Article Text


Original article
Robust pro-inflammatory immune response is associated with serological cure in patients with syphilis: an observational study
  1. Maciej Pastuszczak1,
  2. Anna Gozdzialska2,
  3. Bogdan Jakiela3,
  4. Aleksander Obtulowicz1,
  5. Jerzy Jaskiewicz2,
  6. Anna Wojas-Pelc1
  1. 1Department of Dermatology, Jagiellonian University School of Medicine, Cracow, Poland
  2. 2Andrzej Frycz Modrzewski Cracow University, Cracow, Poland
  3. 3Department of Medicine, Jagiellonian University School of Medicine, Cracow, Poland
  1. Correspondence to Dr Maciej Pastuszczak, Department of Dermatology, Jagiellonian University School of Medicine, Skawinska 8 St., Cracow 31-066, Poland; mpastuszczak{at}


Objectives Approximately 15% of adequately treated patients with early syphilis remain serofast. Pathogenesis and clinical significance of this phenomenon is unclear. The objective of this study was to determine whether there is any association between host immune response and treatment outcome (serofast state or proper serological response).

Methods Forty-four patients with secondary syphilis were enrolled to this study. Levels of pro-inflammatory cytokines such as interferon-γ, tumour necrosis factor-α and interleukin-6 were measured before treatment and 8 hours after injection of antibiotic.

Results After 1 year, based on the serological response patients were stratified into two groups: (1) proper serological response (n=31) and (2) serofast state (n=9). The serological cure rate was 77.5% at 12 months after treatment. Patients with proper serological response had significantly higher levels of analysed cytokines (at baseline and 8 hours after treatment) compared with the serofast state group (p<0.05).

Conclusions We showed that robust host pro-inflammatory immune response to infection may be the predictive factor of serological cure. The treatment outcome may be also associated with the magnitude of immune reaction occurring during the treatment.


Statistics from


Syphilis is an STI caused by Treponema pallidum spp. pallidum. With 12 million new cases a year worldwide, the disease seems to be still a global public health problem. Parenteral penicillin is the therapy of choice at every stage of syphilis, but due to inability to culture T. pallidum in vitro, monitoring of treatment efficacy relies almost entirely on the measurement of immune response (ie, serological assays) rather than direct T. pallidum tests. Appropriate serological response in syphilis has been defined by the Centres for Disease Control and Prevention (CDC) as fourfold or greater decline in the titre of non-treponemal assay when occurred 6 months after beginning of the treatment.1 However, in approximately 15% of patients with early syphilis, the titre of non-treponemal assay during follow-up testing does not meet the criteria of proper response (ie, greater than fourfold decline) and does not decrease considerably, without any clinical evidence of treatment failure or reinfection. Such condition is referred as ‘serofast state syphilis’. Until now, the pathogenesis of serofast state is unknown.2 ,3 Moreover, the absence of novel methods to confirm eradication of the T. pallidum causes uncertainty regarding whether ‘serofast state syphilis’ is a signature of persistent infection or simply a residual immune response in the absence of viable pathogen. Thus, a majority of serofast patients undergo additional treatment despite lack of any evidence to support this practice.

Cell-mediated immunity is thought to be of critical importance during the natural course of T. pallidum infection. T cells participate in the eradication of treponemas primarily through the production of cytokines such as interferon (IFN)-γ, therefore activating macrophages, which subsequently kill opsonised spirochetes.4 ,5 It has been shown recently that suppression of pro-inflammatory (IFN-γ) and concomitant activation of regulatory immune responses (interleukin (IL)-10) may facilitate the progression of syphilis towards later stages of the disease.6 ,7 Hence, the robust pro-inflammatory immune response seems to be crucial to get control over T. pallidum during early infection.

Recent studies suggest that the outcome of syphilis treatment is likely affected by the host immune response. Several factors, which have been implicated as the predictors of appropriate serological response to treatment (ie, younger age, male sex, higher baseline non-treponemal antibody titres and appearance of Jarisch-Herxheimer reaction), seem to be influenced by the host immune response to infection. Interestingly, some studies showed that the appearance of Jarisch-Herxheimer reaction is associated with serological cure in patient with syphilis. During this reaction, there is an increase in inflammatory cytokines such as tumour necrosis factor (TNF)-α and IL-6.8–12 Thus, we hypothesised that baseline levels of immune markers or their change after antibiotic therapy may predict serological cure or serofast state. To explain this, we compared levels of pro-inflammatory cytokines IFN-γ, TNF-α and IL-6 (measured before treatment and 8 hours after injection of penicillin) in patients with syphilis with confirmed serofast state or with proper serological response.


Characteristics of the patients and study design

Forty-four patients (all males, aged 18–55) with confirmed diagnosis of secondary syphilis and no history of syphilis were enrolled at the outpatient clinic of the Department of Dermatology (Jagiellonian University Medical College, Cracow, Poland). Patients were not treated with antibiotics or immunosuppressive agents during preceding 6 months, other chronic inflammatory disorders (eg, autoimmunity) and HIV infection were also excluded. All study participants were positive for serum T. pallidum hemagglutination assay and rapid plasma reagin (RPR) tests. Syphilis staging was determined by a specialist in dermatology and venerology, according to CDC recommendations,1 based on clinical history, physical examination (macular exanthem was found in all subjects) and laboratory findings.

All individuals were given a single intramuscular injection of benzathine penicillin (2.4 IU), followed by 12-hour clinical observation for symptoms of Jarisch-Herxheimer reaction (ie, fever, chills, headache and exacerbation of skin lesions). RPR testing was repeated at 3-month, 6-month and 12-month follow-up visits according to published standards.1 Serological cure was defined as either negative RPR test results or greater than fourfold decrease in the titre at defined time point (ie, 3, 6 or 12 months). Thirteen patients fulfilled criteria of serofast state, defined as either no change in RPR titre or less than twofold decrease when analysed 12 months after treatment. In all subjects with inadequate serological response, lumbar puncture was performed and four patients with increased cerebrospinal fluid (CSF) pleocytosis (≥5 leucocytes/µL) were excluded from further studies. In final analysis, patients were stratified into two groups: (1) those with proper serological response after treatment (serological-cured group, n=31) and (2) patients with improper response (serofast state group, n=9). The epidemiological and clinical characteristics of studied groups are shown in table 1.

Table 1

Characteristics of patients

Cytokine measurements

Blood samples for routine tests, serology and cytokine measurements were taken at baseline (ie, before administration of penicillin) and 8 hours after injection of penicillin. Aliquots of serum were stored at −80°C until analysis. Serum cytokine levels were measured using a high-sensitive ELISA (IFN-γ: R&D Systems, Minneapolis, Minnesota, USA; TNF-α and IL-6: eBiosciences, San Diego, California, USA), according to the manufacturer's instructions. The lower detection limits were as follows: IFN-γ, 0.08 pg/mL; TNF-α, 0.11 pg/mL and IL-6, 0.04 pg/mL. Samples below the threshold of quantification were assigned the value of lower limit of detection.

Statistical analysis

Statistical analysis was performed with the Statistica V.7.1 PL package (StatSoft, 2005). Data are expressed as median and IQR if not otherwise stated. Between-group comparisons were performed with the Mann-Whitney U test and Wilcoxon signed-rank test. p Value <0.05 was considered statistically significant.


Clinical and serological response to therapy

Clinical symptoms of syphilis resolved after therapy in all patients studied. However, the serological cure rate was 67.5% (27/40) at 3 months, 72.5% (29/40) at 6 months and improved to 77.5% (31/40) at 12 months after treatment. In all patients with improper serological response at 12 months after treatment, lumbar puncture was performed. Also, 4 of 13 individuals (30.7%) had ≥5 white blood cells/µL in CSF and protein level in CSF of ≥45 mg/dL. One patient had reactive Venereal Disease Research Laboratory (VDRL) in CSF. None of these patients had neurological symptoms. Nine individuals (22.5%) were classified as serofast. And, 9 of 40 patients had seroreverted for RPR test at 12 months after treatment (ie, had a non-reactive RPR test).

Patients with proper serological response (serological-cured group) had higher baseline RPR titres and higher baseline C-reactive protein (CRP) levels compared with serofast state group (p<0.05; table 1).

Three patients (30%) from the serofast state group and seven (23%) from the serological-cured group experienced exacerbation of skin lesions after first injection of penicillin (difference not statistically significant). Additionally, rise of body temperature was reported in most patients up to 8 hours after first injection of drug, with no significant difference in magnitude of the response between the two studied groups (table 1).

Differences in cytokine levels in patients with serologically cured and serofast syphilis

Patients with proper serological response had significantly higher baseline levels of IFN-γ, TNF-α and IL-6 compared with the serofast state group (p<0.05; figure 1).

Figure 1

Levels of analysed cytokines before and after treatment. IFN-γ, interferon-γ; IL-6, interleukin-6; TNF-α, tumor necrosis factor-α.

In serological-cured patients, there was a significant decrease in serum levels of all measured cytokines 8 hours after first injection of penicillin (p=0.02). However, in the serofast state group serum IFN-γ remained unchanged (figure 1A), while levels of TNF-α and IL-6 decreased (figure 1B and C, respectively).

Eight hours after first injection of penicillin, patients with proper serological response had still higher serum levels of IFN-γ, TNF-α and IL-6 compared with the serofast state group (figure 1). However, the difference for IL-6 was not statistically significant (figure 1).


Most patients achieve serological cure after syphilis treatment. However, in a substantial proportion (15–20%) of patients, non-treponemal titres neither increase nor decrease fourfold after treatment and such patients are referred as serofast.3 Clinical significance and pathomechanism of serofast state is unknown. Thus, the optimal management of this particular group of patients with syphilis is unclear. In our study, the prevalence of serofast state was 29.5% and is in accordance with other studies.2 ,10 ,11

In a substantial proportion of patients with syphilis, the symptoms of disease diminish despite lack of treatment.13 ,14 Cell-mediated immunity to infection is believed to be crucial for this process. IFN-γ-activated macrophages seems to be the most important component of the immune system involved in the eradication of T. pallidum.5 Moreover, it has been shown that robust pro-inflammatory response at early stages of infection may decrease the risk of progression to later stages of syphilis.6 Hence, the question arose whether serofast state could be associated with a lower magnitude of innate immune response.

Several studies found that individuals with higher baseline titres of non-treponemal tests were more likely to achieve serological cure.8–11 This phenomenon was partially explained by Baker-Zander et al,15 who demonstrated that VDRL-immunised rabbits exhibited partial protection against reinfection with T. pallidum. These investigators hypothesised that high antibody titres of VDRL may facilitate the clearance of pathogen. If so, higher baseline titres of non-treponemal assays may signify a beneficial immune response to T. pallidum.

Our data show that in patients with secondary syphilis, higher baseline non-treponemal titres are correlated with higher levels of pro-inflammatory cytokines such as IFN-γ, TNF-α and IL-6. Moreover, we found that higher baseline levels of measured cytokines were associated with favourable outcome, that is, serological cure at 12 months after therapy. Thus, it seems that robust host pro-inflammatory immune response to infection may be the predictive factor of serological cure.

Bactericidal action of penicillin causes the release of treponemal lipoproteins that may easily activate immune response mainly via Toll-like and other pattern recognition receptors.16 As a result of this, some patients may experience short-term constitutional symptoms such as fever, chills, headache and exacerbation of existing cutaneous lesions. These symptoms result from transient elevation of pyrogenic cytokines, mainly TNF-α and IL-6. Treponemal lipoproteins released from damaged or killed pathogens are thought to be the triggers of this reaction.17 Of interest, in our study we found that serological-cured patients had higher levels of pro-inflammatory cytokines that were measured shortly after injection of penicillin. Thus, the treatment outcome may be related not only with host immune response to infection, but also with the magnitude of immune reaction occurring during the treatment.

Until now, it is unclear why some patients develop robust and some weaker immune response to treponemal infection and during the treatment. Presumably, functional single-nucleotide polymorphisms for genes coding pro-inflammatory cytokines and genetic variability within syphilis strains may be relevant.18 ,19

Most patients in our study achieved serological cure at 3 months after completing the treatment. In the next months, serological cure rate improved only slightly. These findings generally support those of other investigations regarding that the likelihood of a substantial decline in the serofast proportion over time is low.8–11

In our study, four patients with improper serological response at 12 months after therapy had asymptomatic neurosyphilis. Until now, only three studies performed lumbar punctures among individuals with serological non-response to therapy. None of the evaluated patients were found to have reactive CSF suggestive of neurosyphilis in the study of Fiumara;20 however, as far as 17 and 115 of 205 individuals were identified in the study of Zhou et al21 and Li and Zhang,22 respectively. Further studies are needed to establish the association between serofast state and the risk of neurosyphilis. However, lumbar puncture should be considered in patients with serofast state syphilis.

Our study has some limitations. Given the small sample size of this study and patients with secondary syphilis included, we acknowledge that our data need to be interpreted cautiously. Despite increasing number of cases worldwide, syphilis remains a relatively rare disease in developed countries. As secondary syphilis is a result of T. pallidum bacteraemia, the immune response at this stage may be more pronounced. Second, in order to address the question of persisting spirochetes after therapy in serofast individuals, highly sensitive direct detection methods for T. pallidum would be required. However, it is not possible to cultivate T. pallidum in vitro and PCR-based assays are characterised by poor sensitivity and specificity. Thus, it still remains unclear whether serofast state reflects persistent T. pallidum or an altered immune process in which antibodies targeting T. pallidum phospholipids are not downregulated. Third, the invasion of T. pallidum to the central nervous system may occur even at an early stage of syphilis. Asymptomatic neurosyphilis may be the factor affecting the treatment outcome. However, in the current study, a CSF examination prior to the treatment was not performed. Finally, the bacterial lysis caused by penicillin leads to increase in levels of pro-inflammatory cytokines. In contrast, we found decrease in levels of cytokines 8 hours after penicillin injection in all included patients. However, the differences between groups were still pronounced. Presumably, the time point for analysis should be set earlier.

In conclusion, our data support the hypothesis that baseline pro-inflammatory immune response to treponemal infection may be an important predictor of treatment outcome. In support of this, we identified that robust pro-inflammatory response occurring shortly after treatment may be the prognostic factor of serological cure in patients with syphilis. However, further studies on larger groups are needed in order to confirm our observations and evaluate the role of immune system in pathomechanism of serofast state.

Key messages

  • Improper serological response after adequate syphilis therapy may be associated with neurosyphilis.

  • Even every fifth adequately treated patient with early syphilis may remain serofast.

  • Robust host pro-inflammatory immune response to infection may be the predictive factor of serological cure in patients with syphilis.

  • Further studies are needed to ascertain the host factors that determine the magnitude of immune response to treponemal infection.


View Abstract
  • Abstract in Polish

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Jackie A Cassell

  • Contributors MP, BJ and AO contributed to the conception and design of the study, wrote and drafted the manuscript and were involved in critical revision of the manuscript. AG and JJ processed and analysed the samples, helped with statistical analysis and reviewed the manuscript. AW-P contributed to the conception of the study and was involved in critical revision of the manuscript.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Jagiellonian University Bioethics Committee (approval number KBET/164/B).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.