The Diagnosis of the French Disease in the Year 2020

The spread of syphilis in 1494/1495 in Italy is the first known instance of an outbreak of the disease. It is believed the disease was brought in from the New World by Columbus and his crew members and following the invasion of France in Naples, Italy, the disease began to spread throughout Europe. Known as the “French Disease” (because the French brought it in to Italy), it has had a significant impact that transcended medicine and had societal and historical implications. Despite its long history, we have no clue how to best diagnose this beast. Most of this stems from the fact that syphilis tends to be difficult to grow in the microbiology lab. Like another spirochete, it has a narrow temperature and environment at which it will grow in the lab. As such, we have relied on indirect methods of diagnosis, which include antiphospholipid antibodies and more specific treponemal antibodies. This post will serve as a review of the diagnostic modalities that we have in 2020.

Direct Methods:

Before going into the more well-known methods such as non-treponemal antibodies, there is a body of literature evaluating direct methods of diagnosis. Most of these, however, are not available to most labs and/or require expertise. A recent review (1) evaluated many of these methods, including dark field microscopy, direct fluorescence antibody, silver staining, and nucleic acid amplification testing in various stages of syphilis. The main highlights are below:

Notably, DFA seems to be the best for primary syphilis, using pAbs as its target when compared to dark-field microscopy. Unfortunately, a lot of these assays are not available for commercial use. They do have the advantage over DFM in that you do not need to get the sample to a lab ASAP.

Before going forward, we probably need to ask what is dark-field microscopy? Whenever looking at bacteria under the microscope, in an ideal situation, we would like to stain it to be able to contrast it from the background:

Syphilis is difficult to culture and isolate, in general. It is also very difficult to stain because it is so thin, so one way to look at it is to see its movement and morphology. To do this, you need to keep it alive to see it “vibe.” But you need to contrast it from the background to have a better look. So how do you do this? Dark-field microscopy. The principle here is to get the light to reflect off the specimen, without it picking up any of the background. This allows the bacteria to “light up” while the rest of the background remains dark, hence giving you the needed contrast.

There are a few caveats. Because you need the bacteria to remain alive to see its movement, results depend on rapid access (within 20min of getting the sample) as well as technologist proficiency in identifying T. pallidum from other spirochetes that are present in normal flora. Ok, back to direct methods.

Looking at the above table, it seems silver staining and immunohistochemistry are basically not worth it with its low sensitivities. This may be due to the challenges of stain interpretation and the fact that certain melanin and reticulin fibers can mimic the appearance of spirochetes in tissues. Hence, this is unlikely to be encountered in clinical practice.

The most compelling of these methods, however, is nucleic acid amplification testing. Samples can be obtained from different sources and can be used at different stages. Most of these NAATs use tpp47 and polA as their targets and are very specific. Their sensitivity, however is not terribly great for a screening test, with the sensitivity dropping in secondary syphilis:

Further, these assays tend to have low sensitivity if using whole blood or blood fractions such as serum, plasma, or peripheral blood mononuclear cells. For instance, many assays have a sensitivity ranging from 0 to 55% using tpp47 and 0 to 62% using polA NAATs. Despite the data available, a lot of these nucleic acid amplification methods are not available for commercial use.

TL;DR – For direct methods of detection, DFA is the most sensitive and specific. This is followed by Dark field microscopy, but it requires an actual dark field microscope and someone who knows what they are doing. NAAT would be up next, but its sensitivity in secondary syphilis is lacking, plus it is not commercially available. Do not use immunohistochemistry here.

Non-Treponemal Antibodies

These are essentially “antiphospholipid antibodies” that are made in response to phosphatidylcholine taken up by Syphilis. There are 2 of these, including the rapid plasma regain (RPR) and the venereal disease research laboratory (VDRL) antibodies. Both are typically easier to run than the older treponemal antibodies that were used to confirm syphilis (more on that later), hence “back in the day” these were useful for screening purposes. Overall, RPR tends to be more sensitive in most cases, with the exception of CSF, where VDRL tends to be more sensitive. In general, however, both can be use interchangeably.

In primary syphilis, both VDRL and RPR tend to have a sensitivity of 62-78% for diagnosis when compared to dark field microscopy, with some studies quoting a sensitivity of 92.7% for RPR, and another study quoting a sensitivity of 90.6% for RPR for HIV patients with genital ulcers (in comparison to 87.3% for those without HIV), though the sensitivity was compared to a multiplex PCR for syphilis, which is not available commercially:

The diagnostic sensitivity of non-treponemal antibodies increases in secondary syphilis (again, makes sense since there is more time for antibodies to develop). The sensitivities for both assays are near 100%:

Data for early latent is mostly on VDRL and it does well here, with a sensitivity of 85-100%. In late latent syphilis, however, the sensitivity for both RPR and VDRL drops to 64-75% and 61-64%, respectively.

In looking at data for those who had unknown stages of syphilis, overall RPR performed better than VDRL.

Why is this? It is unclear, but there is data that supports that, even in those who are not treated for syphilis, the proportion of patients testing positive for syphilis drops significantly:

Non-treponemal antibodies are also used for the diagnosis of neurosyphilis. In general, it is difficult to figure out how these test perform, since in many studies, CSF VDLR or RPR positivity is used as a marker of disease. Regardless, it seems that CSF RPR is less sensitive than CSF VDRL, with CSF RPR having a quoted sensitivity of 51-82%. The latter has a sensitivity range of 48-87.5%, when compared to CSF PCR, symptoms (usually new vision or hearing loss), positive CSF particle agglutination assay, or CSF WBC >10 or protein >45.

For ocular syphilis, using CSF, the sensitivity of CSF VDRL is usually <50%.

False-negativity of Non-treponemal Assay

This is usually attributed to the prozone effect. This occurs in assays that use antibodies, such as the sandwich ELISA. There needs to be a balance between antigen and antibody to allow the “sandwiching” to occur (is that a word? I’ll run with it). If there is too much antigen, then all antibody sites will be occupied and there won’t be any chance for the capturing antibodies to detect the antigen (aka there is no “sandwiching”). On the flip side, if there is too much blocking antibody (aka not the antibody that is part of the reagent), then you won’t pick up the “sandwiching” either.

The quoted rate of false negativity is <0.85%. Usually, this is seen in primary and secondary syphilis, but can be seen in pregnancy.

False Positivity

Because these are non-treponemal antibodies, more specifically, antiphospholipid antibodies, it is not unexpected that there is cross-reactivity with other diseases. The false positivity rate tends to be less than 1.5% and it is defined in certain autoimmune disease, leprosy, yaws (another treponema) and HIV. There is some limited data on its association with malaria, hep B, hep C, and IV drug abuse. Regardless, the titers here tend to be <1:8.

Treponemal Tests

Prior to the automated test we have today, we had 2 more labor intensive tests for syphilis (again, indirect methods here). These were the hemagglutination method and the fluorescent treponemal antibody-absorbent method.

The FTA-ABS essentially measures antibody in a round-about way:

1. Get patient’s serum and mix it with a solution to remove non-specific antibodies (aka non-syphilitic antibodies). This is the absorbent part of the equation
2. Once that is done, put the patient’s serum into a slide with a suspension of T.pallidum that is obtained from infecting a poor rabbits testicle (usually the left testicle). Remember, T.pallidium is hard to culture, so we need a live specimen to grow it in.
3. Once this is done, the patient’s antibodies will coat the syphilis in the slide. Then add the fluorescent antibodies
4. Watch as they glow:

The hemagglutination has 2 variations: the treponemal pallidum particle agglutination (TP-PA) and the microhemmaglutination assay for T.pallidum antibodies (MHA-TPA). Both work under the same principle, in which the patient’s antibody will aggregate (or agglutinate) a bunch of RBCs. How is this done?

1. RBCs are sensitized with ultrasonicated material from T. pallidum
2. Patients serum is added to the RBCs
3. Agglutination occurs

You can see these tend to be a bit labor intensive. There is quite a bit of work, when compared to measuring non-treponemal antibodies. Hence, for the “Traditional algorithm” the non-treponemal antibodies are used for screening and the more labor-intensive specific tests are then used for confirmation. Now on to the data (3):

Looking at the manual assays, all seem to have roughly the same sensitivity beyond primary syphilis. MHA-TP has the lowest sensitivity for primary syphilis, followed by FTA-Abs and then TP-PA. All seem to perform well at later stages, which is not surprising since these tend to be positive for life in many patients:

For neurosyphilis, the sensitivity of CSF FTA-ABS was 90.9-100% when comparing it to CSF VDRL, but it dropped to anywhere from 22.2-100% when using clinical signs and symptoms. Specificity for it ranged from 55-100%. CSF TP-PA had a sensitivity of 75.6% to 95% with specificity ranging from 85.5 to 100%, depending on the study.

Newer assays are typically enzyme immunoassays, chemiluminescent immunoassay, or ELISA. In general, these have comparable sensitivities to the manual assays for secondary syphilis and beyond. For primary syphilis, the Captia syphilis-G assay (an EIA) and the Trep-sure assay had the lowest sensitivities, 82.3% and 53.8% respectively:

TL;DR – manual assays and immunoassays have a similar sensitivities for anything beyond secondary syphilis. For primary syphilis, most immunoassays beat the manual assays in terms of sensitivity. For anything beyond primary syphilis, all perform fantastic.

Reverse Algorithm Vs Traditional Algorithms:

Given the availability of automated treponemal antibody tests, more labs are moving towards the reverse algorithm since it is more efficient and requires less labor. Labs that have high volumes of tests tend to gravitate towards using the reverse algorithm. Which one is used depends on one’s particular institution and it may not be significantly different. For instance, one retrospective study (5) evaluated both algorithms in an academic medical center found that the reverse algorithm had more positive screens and true positives (0.7% vs 0.4% p-value 0.002), however the false positive rates were not significantly different. One advantage of the reverse algorithm is that it can catch early cases of syphilis (primary) and those with late latent syphilis who may have been missed by the traditional algorithm (6).

Because of this, using the reverse algorithm, you can come across discordant results. Remember, that treponemal antibodies tend to be forever while non-treponemal antibodies can become undetectable after some time (see above). If this happens, a few things you need to consider:

1. Someone has been treated before for syphilis
2. They are early on into their infections (so the non-treponemal antibodies are not there yet)
3. They are very late latent (i.e. their non-treponemal antibodies are not there anymore).

If there is discordance, then using a treponemal test as a tiebreaker is reasonable (preferably a different one from the one that was used). If the repeat is positive, then treat for syphilis. If negative, then it can be assumed that it is a false positive and do not treat. This is based (7) on a cross-sectional analysis of over 21,000 patients where the reverse algorithm was implemented. In 255 discordant cases, 71 tested negative for TP-PA. follow up of this group revealed that only one patient, a young HIV positive asymptomatic male, seroconverted their TP-PA to positive. This suggests that in a low risk population, a negative treponemal test can be used as a tie breaker in cases of discordance.

What to do with an indeterminate TP-PA?

This is unclear. In one study (8), of 6 indeterminate TP-PA, only one resulted positive when tested with another treponemal assay. Further, another study (9) using electrochemiluminescence in discordant results, the cutoff index was closer to the negative TP-PA than to a positive one:

In these instances, it may be reasonable to treat if someone is a high risk patient (HIV positive, MSM, high risk behaviors) and repeat with another treponemal assay if low risk.

Point of Care Syphilis Testing

The Syphilis Health Check is a rapid qualitative test for detection of antibodies to T. pallidum in serum, plasma, or whole blood. This is a lateral flow assay, that works similarly to a pregnancy test:

Overall (10), the sensitivity ranged from 50-89% in clinical studies vs 77.8% to 100% in regulatory studies. Fingerstick tended to have lower sensitivities, except for one regulatory study. The pooled sensitivity for these prospective studies was 87.7% (95% CI 71.8%-97.2%) and specificity was 96.7% (95% CI 91.9%-99.2%):

Overall, using treponemal test as reference, the overall sensitivity was fairly high:

Sensitivity dropped when using treponemal and non-treponemal tests as reference:

Why was there such a discrepancy between studies? The authors cite this is likely due to the fact that FDA trials then to include rigorous training and oversight, while trial studies do not include methods of quality control. How true this is, I am not sure. Overall, seems that whole blood samples tend to do well, and finger-sick is actually a viable option for this. The only issue is it does not discriminate between past and current infection, and you still need a non-treponemal assay to trend for treatment purposes.

References:

1. Elitza S Theel, Samantha S Katz, Allan Pillay, Molecular and Direct Detection Tests for Treponema pallidum Subspecies pallidum: A Review of the Literature, 1964–2017, Clinical Infectious Diseases, Volume 71, Issue Supplement_1, 1 July 2020, Pages S4–S12, https://doi.org/10.1093/cid/ciaa176
2. Susan Tuddenham, Samantha S Katz, Khalil G Ghanem, Syphilis Laboratory Guidelines: Performance Characteristics of Nontreponemal Antibody Tests, Clinical Infectious Diseases, Volume 71, Issue Supplement_1, 1 July 2020, Pages S21–S42, https://doi.org/10.1093/cid/ciaa306
3. Rudolph AH. The microhemagglutination assay for Treponema pallidum antibodies (MHA-TP), a new treponemal test for syphilis: where does it fit?. J Am Vener Dis Assoc. 1976;3(1):3-8.
4. Ina U Park, Anthony Tran, Lara Pereira, Yetunde Fakile, Sensitivity and Specificity of Treponemal-specific Tests for the Diagnosis of Syphilis, Clinical Infectious Diseases, Volume 71, Issue Supplement_1, 1 July 2020, Pages S13–S20, https://doi.org/10.1093/cid/ciaa349
5. Dunseth, Craig D et al. “Traditional versus reverse syphilis algorithms: A comparison at a large academic medical center.” Practical laboratory medicine vol. 8 52-59. 29 Apr. 2017, doi:10.1016/j.plabm.2017.04.007
6. Park IU, Chow JM, Bolan G, Stanley M, Shieh J, Schapiro JM. Screening for syphilis with the treponemal immunoassay: analysis of discordant serology results and implications for clinical management. J Infect Dis. 2011;204(9):1297-1304. doi:10.1093/infdis/jir524
7. Lin JS, Eder M, Bean S. Screening for Syphilis Infection in Pregnant Women: A Reaffirmation Evidence Update for the U.S. Preventive Services Task Force [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2018 Sep. (Evidence Synthesis, No. 167.) Figure 1, Syphilis Serologic Screening Algorithms* Available from: https://www.ncbi.nlm.nih.gov/books/NBK525911/figure/ch1.fig1/
8. Lee, Kyunghoon et al. “Characterization of sera with discordant results from reverse sequence screening for syphilis.” BioMed research international vol. 2013 (2013): 269347. doi:10.1155/2013/269347
9. Lee S, Yu HJ, Lim S, Park H, Kwon MJ, Woo HY. Evaluation of the Elecsys Syphilis electrochemiluminescence immunoassay as a first-line screening test in the reverse algorithms for syphilis serodiagnosis. Int J Infect Dis. 2019;80:98-104. doi:10.1016/j.ijid.2018.12.016
10. Bristow CC, Klausner JD, Tran A. Clinical Test Performance of a Rapid Point-of-Care Syphilis Treponemal Antibody Test: A Systematic Review and Meta-analysis. Clin Infect Dis. 2020;71(Supplement_1):S52-S57. doi:10.1093/cid/ciaa350