Syphilis Bacteria Found in Spinal Fluid Across All Disease Stages

The bacterium that causes syphilis has long been known to invade the central nervous system, but a new study suggests that this invasion may be far more common, and far less tied to obvious neurological symptoms, than many clinicians assume. In research published in BMC Infectious Diseases, a team at Shanghai Skin Disease Hospital affiliated with Tongji University detected genetic material from Treponema pallidum in the cerebrospinal fluid of patients at every stage of syphilis, including early primary infections and asymptomatic latent disease. The finding adds weight to a growing body of evidence that the spirochete responsible for syphilis reaches the brain and spinal cord more often than standard laboratory tests reveal, complicating long-standing assumptions about when and how the infection should be treated.

The study enrolled 155 untreated patients with confirmed syphilis who were all negative for HIV, a deliberate design choice that removes one of the most confounding variables in this field. Much of what is known about neurosyphilis comes from studies of people coinfected with HIV, whose immune status can alter both the course of syphilis and the behavior of the spirochete within the nervous system. By restricting the cohort to HIV-negative individuals, the researchers aimed to characterize the baseline frequency of central nervous system invasion in an otherwise immunocompetent population. The participants fell into three groups: 47 with primary or secondary syphilis, 70 with latent syphilis, and 38 with symptomatic neurosyphilis, the stage at which the infection has already produced recognizable neurological disease.

From each patient, the team collected paired samples of cerebrospinal fluid, drawn by lumbar puncture, and blood plasma. They then searched both compartments for T. pallidum DNA using a nested polymerase chain reaction assay, a highly sensitive amplification technique that targets two independent genetic markers: the polA gene, which encodes DNA polymerase I, and the Tpp47 gene, which encodes a 47-kilodalton membrane protein. Requiring amplification of two separate targets reduces the risk of false positives from contamination or non-specific amplification, a persistent concern when testing biological fluids for organisms that are present at very low copy numbers. Nested PCR, in which a second round of amplification is performed on the products of the first, pushes sensitivity further, allowing detection of even trace amounts of bacterial DNA.

The results were striking in their consistency across disease stages. T. pallidum DNA was detected in the cerebrospinal fluid of 17.0 percent of patients with primary or secondary syphilis, 17.1 percent of those with latent syphilis, and 28.9 percent of those with symptomatic neurosyphilis. The differences among the three groups were not statistically significant, with a P value of 0.284, meaning the observed variation could plausibly reflect chance. In practical terms, roughly one in six patients with early or latent disease, and nearly one in three with symptomatic neurosyphilis, carried detectable spirochete DNA in the fluid bathing their brain and spinal cord. The similarity between the early and latent groups suggests that central nervous system invasion is not a late complication reserved for advanced disease but may occur soon after initial infection and persist silently.

Perhaps the most provocative finding emerged when the researchers examined patients whose routine cerebrospinal fluid tests were entirely unremarkable. Among individuals with negative CSF results on both the treponemal particle agglutination assay, known as TPPA, and the rapid plasma reagin test, RPR, together with a normal white blood cell count and normal protein concentration in the spinal fluid, 14.6 percent nonetheless tested positive for T. pallidum DNA. That subgroup included three patients with primary or secondary syphilis and three with latent syphilis. These are patients who, under conventional diagnostic algorithms, would be considered to have no laboratory evidence of central nervous system involvement, yet molecular testing revealed bacterial genetic material in their spinal fluid.

This discrepancy exposes a fundamental limitation of the tools clinicians currently rely on. The diagnosis of neurosyphilis has never been straightforward: there is no single gold-standard test, and guidelines typically combine CSF white cell counts, protein levels, and non-treponemal serology performed on spinal fluid. Those markers reflect inflammation and antibody production rather than the presence of the organism itself. A patient can harbor spirochete DNA without mounting the inflammatory response that would elevate the white cell count, and antibody tests on CSF can be negative even when the bacterium is present. Molecular detection offers a more direct window, though it too carries caveats, since DNA fragments can persist after organisms have died and the clinical meaning of a positive PCR result in an asymptomatic patient remains uncertain.

The study also compared detection in spinal fluid and blood plasma, and the agreement between the two compartments was poor, with a kappa statistic of just 0.1 and a 95 percent confidence interval of 0.0 to 0.3. A kappa value near zero indicates agreement barely better than chance, meaning a positive plasma result told clinicians almost nothing about whether the organism could be found in the spinal fluid, and vice versa. This finding matters because it argues against the idea that CSF positivity is simply a passive spillover of bacteria circulating in the bloodstream. If blood-borne organisms were passively contaminating the spinal fluid, the two measurements would be expected to track together more closely. The discordance instead supports the interpretation that T. pallidum invades and establishes itself within the central nervous system as a distinct event, independent of the level of bacteria in peripheral blood.

When the team ran a multivariable analysis to identify which clinical and laboratory features independently predicted a positive CSF PCR result, one factor stood out: an elevated white blood cell count in the spinal fluid, defined as more than eight cells per microliter. Patients with this degree of CSF pleocytosis had roughly seven times the odds of testing positive for T. pallidum DNA compared with those whose counts were normal, with an adjusted odds ratio of 7.18 and a 95 percent confidence interval of 1.86 to 27.68, reaching statistical significance at P equals 0.004. This association makes biological sense, since an influx of white cells into the spinal fluid signals active inflammation, which is more likely when living organisms are present and provoking an immune response. Yet the finding cuts both ways: because a substantial minority of patients with completely normal CSF parameters still had detectable DNA, inflammation alone cannot serve as a reliable gatekeeper for deciding who harbors the organism.

The authors are careful to frame their conclusions with appropriate caution. Detecting bacterial DNA does not prove the presence of viable, replicating spirochetes, and the clinical significance of a positive PCR result in a patient without neurological symptoms or abnormal CSF findings remains unresolved. Whether such patients are at risk of progressing to symptomatic neurosyphilis, whether they require treatment regimens designed to achieve high drug concentrations in the central nervous system, and whether the DNA signal clears spontaneously are all questions the study cannot answer. The cross-sectional design captures a single moment in time for each patient, so prospective follow-up will be needed to determine what CSF PCR positivity predicts for long-term outcomes.

Even with those caveats, the study carries practical implications for a disease that is resurging worldwide. Syphilis cases have climbed steadily in many countries over the past decade, and clinicians continue to debate which patients warrant lumbar puncture and whether standard benzathine penicillin regimens are adequate when the nervous system is involved. The demonstration that roughly one in six HIV-negative patients with early or latent syphilis shows molecular evidence of central nervous system invasion, often without any abnormality in conventional CSF tests, challenges the comfort of current algorithms. It suggests that the spirochete’s journey into the brain and spinal cord is an early and frequent feature of infection rather than a rare endpoint, and that the true burden of neurological involvement may be systematically underestimated. As molecular diagnostics become more widely available, studies like this one will help determine whether PCR testing of spinal fluid should move from the research bench into routine clinical practice, and how a positive result should reshape treatment decisions for patients at every stage of this ancient and still-elusive infection.

Subject of Research: Detection of Treponema pallidum DNA in cerebrospinal fluid across all stages of syphilis in HIV-negative patients

Article Title: Detection of Treponema pallidum DNA in cerebrospinal fluid and plasma across all stages of syphilis in HIV-negative patients: prevalence and clinical correlates

Article References: Wang, C., Shi, M., Zou, D., Chen, Z., & Long, F. (2026). Detection of Treponema pallidum DNA in cerebrospinal fluid and plasma across all stages of syphilis in HIV-negative patients: prevalence and clinical correlates. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14534-6

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14534-6

Keywords: Treponema pallidum, syphilis, neurosyphilis, cerebrospinal fluid, nested PCR, HIV-negative patients, central nervous system invasion, CSF white blood cell count, latent syphilis, molecular diagnostics, BMC Infectious Diseases, spirochete