Neurosyphilis

INTRODUCTION — The term "neurosyphilis" refers to infection of the central nervous system (CNS) by Treponema pallidum, subspecies pallidum (hereafter termed T. pallidum). Neurosyphilis can occur at any time after initial infection.

Early in the course of syphilis, the most common forms of neurosyphilis involve the cerebrospinal fluid, meninges, and vasculature (asymptomatic meningitis, symptomatic meningitis, and meningovascular disease). Late in disease, the most common forms involve the brain and spinal cord parenchyma (general paralysis of the insane and tabes dorsalis). Each form has characteristic clinical findings, but in some cases there is overlap between these findings.

This topic will review the pathogenesis, epidemiology, clinical findings, diagnosis, and treatment of neurosyphilis. Other aspects of syphilis are discussed separately. (See "Syphilis: Epidemiology, pathophysiology, and clinical manifestations in patients without HIV" and "Syphilis: Treatment and monitoring" and "Syphilis in patients with HIV".)

PATHOGENESIS — Neurosyphilis begins with invasion of the cerebrospinal fluid (CSF), a process that probably occurs shortly after acquisition of T. pallidum infection. The organism can be identified in the CSF from approximately one-quarter of untreated patients with early syphilis [1,2]. Specific strains of T. pallidum may be more likely to cause neurosyphilis [3].

Unlike other bacteria that can infect the CSF, invasion of CSF with T. pallidum does not always result in persistent infection, as spontaneous resolution may occur in some cases without an inflammatory response. In other cases, spontaneous resolution may occur after a transient meningitis (figure 1).

Persistent meningitis is the result of failure to clear organisms from the CSF. Patients with persistent meningitis have "asymptomatic neurosyphilis," and individuals with this form of neurosyphilis are at risk for subsequent forms of symptomatic neurosyphilis (figure 1) [4]. Studies performed in the early 20^th century showed that the more abnormal the CSF in asymptomatic meningitis, the more likely that symptomatic neurosyphilis would develop [5]. In the antibiotic era, the risk of developing symptomatic neurosyphilis in individuals with asymptomatic neurosyphilis, particularly after antibiotic treatment of uncomplicated syphilis, is unknown. (See 'Spinal fluid examination' below.)

The mechanism of T. pallidum clearance from CSF is probably similar to the immune response that occurs in peripheral infection, where opsonized organisms are cleared by activated macrophages. Supporting evidence comes from a study of nonhuman primates infected with T. pallidum; the number of CSF CD4+ T cells and the amount of gamma-interferon produced by CSF lymphocytes increased throughout the period of bacterial clearance, consistent with a "Th-1-type" cellular immune response [6].

EPIDEMIOLOGY — Neurosyphilis was common in the pre-antibiotic era, occurring in 25 to 35 percent of patients with syphilis [4,7]. Of these, approximately one-third had asymptomatic neurosyphilis, one-third had tabes dorsalis, and at least 10 percent had paresis. Meningovascular syphilis was seen in about 10 percent, with the remaining patients having other forms of neurosyphilis, including symptomatic meningitis and cranial nerve abnormalities.

In the current era, neurosyphilis, particularly the early forms, is most frequently seen in persons with HIV. This association may simply reflect the fact that syphilis is most common in men who have sex with men, many of whom have HIV, or it may reflect a true difference in susceptibility [8]. (See "Syphilis in patients with HIV", section on 'Epidemiology'.)

For persons with HIV, the risk of asymptomatic and symptomatic neurosyphilis is increased in those with lower peripheral CD4+ T cell counts, detectable plasma HIV RNA, and/or no antiretroviral therapy [9-14].

The frequency of the late forms of neurosyphilis (general paresis and tabes dorsalis) has declined in the antibiotic era, with the result that these forms, particularly tabes dorsalis, are now uncommon [15,16]. Many experts believe that this decline is due, at least in part, to the widespread use of antibiotics for unrelated illnesses.

CLINICAL MANIFESTATIONS — Neurosyphilis can be classified into early forms and late forms (figure 1). The early forms typically affect the cerebrospinal fluid (CSF), meninges, and vasculature, while the late forms affect the brain and spinal cord parenchyma. Most of what we know about the clinical manifestations of neurosyphilis comes from observations made before the advent of penicillin [4,7].

Early neurosyphilis

Asymptomatic neurosyphilis — By definition, patients with asymptomatic neurosyphilis have no symptoms or signs of central nervous system disease, although they may have evidence of concomitant primary or secondary syphilis. Asymptomatic neurosyphilis can occur within weeks to months after infection, but less commonly occurs more than two years after infection.

The diagnosis is based on the identification of CSF abnormalities, including a lymphocytic pleocytosis that is typically <100 cells/microL, an elevated protein concentration that is usually <100 mg/dL, a reactive CSF-Venereal Disease Research Laboratory test (VDRL), or a combination of these abnormalities.

In patients with suspected asymptomatic neurosyphilis who do not have HIV infection, a CSF lymphocyte count >5 cells/microL or a protein concentration >45 mg/dL is consistent with the diagnosis of neurosyphilis. Establishing the diagnosis of asymptomatic neurosyphilis in patients who have HIV infection with CSF pleocytosis but nonreactive CSF-VDRL is difficult because mild CSF pleocytosis and elevated protein can be due to HIV itself. (See 'Diagnosis' below.)

Patients with asymptomatic neurosyphilis, regardless of CSF-VDRL reactivity, should be treated for neurosyphilis to prevent progression to symptomatic disease. (See 'Treatment' below.)

Symptomatic meningitis — Most often, symptomatic meningitis occurs within the first year after infection, but it can occur years later. As with asymptomatic neurosyphilis, peripheral findings of early syphilis may coexist, particularly the rash of secondary disease.

Patients with symptomatic syphilitic meningitis complain of headache, confusion, nausea and vomiting, and stiff neck. Visual acuity may be impaired if there is concomitant uveitis, vitreitis, retinitis, or optic neuropathy. Signs include cranial neuropathies, particularly of the optic, facial, or auditory nerves. (See 'Ocular syphilis' below and 'Otosyphilis' below.)

Meningitis may cause hydrocephalus as well as arteritis of small-, medium-, or large-sized vessels, leading to ischemia or infarction of brain or spinal cord. (See 'Meningovascular syphilis' below.)

Focal meningeal inflammation may lead to diffuse leptomeningitis or to syphilitic gummas, which are focal areas of inflammation that present as mass lesions contiguous with the leptomeninges (image 1) [17]. Meningitis, brain ischemia, or gummas may cause seizures.

Syphilitic meningitis may uncommonly affect the spinal cord and cause meningomyelitis or hyperplastic pachymeningitis with polyradiculopathy. Symptoms and signs include back pain, sensory loss, incontinence, leg weakness, or muscle atrophy.

The CSF abnormalities that accompany symptomatic meningitis are more severe than those seen in asymptomatic meningitis. CSF lymphocyte counts are generally between 200 to 400 cells/microL, CSF protein concentration is typically between 100 to 200 mg/dL, and CSF-VDRL is almost always reactive. Neuroimaging may show enhancement of meninges, spinal fluid, cranial nerves, or spinal roots. Cerebral gummas show focal areas of enhancement adjacent to the meninges, often with a "dural tail," and surrounding edema (image 1).

Ocular syphilis — Ocular syphilis can involve almost any eye structure, but posterior uveitis and panuveitis are the most common and present with diminished visual acuity [18]. Additional manifestations may include optic neuropathy, interstitial keratitis, anterior uveitis, and retinal vasculitis. Ocular syphilis is often, but not always, accompanied by syphilitic meningitis. Ocular syphilis may worsen when the diagnosis is not considered and patients are treated with systemic or topical corticosteroids [19]. Reports first published in 2015 suggested that rates of ocular syphilis are increasing in the United States [20,21]. A preliminary study found no single or predominant strain of T. pallidum was causing ocular syphilis [22].

Otosyphilis — Hearing loss, with or without tinnitus, should be considered as part of the neurologic symptoms or signs of neurosyphilis [23,24]. Like ocular syphilis, hearing loss may or may not be accompanied by meningitis.

Meningovascular syphilis — As with other bacterial meningitides, syphilitic meningitis can cause an infectious arteritis that may affect any vessel in the subarachnoid space surrounding the brain or spinal cord and result in thrombosis, ischemia, and infarction. This form of neurosyphilis may present as an ischemic stroke in a young person. Stroke may develop at any time from the first months to the first few years after infection, with an average interval of seven years in a series conducted in the pre-antibiotic era [4].

Many patients with meningovascular syphilis have prodromal symptoms, such as headache, dizziness, or personality changes, for days or weeks before the onset of ischemia or stroke. These symptoms are probably due to concomitant meningitis.

The manifestations of cerebral ischemia may be acute or chronic. The neurological deficits reflect the territory of the vessel involved. The middle cerebral artery and its branches are most commonly affected. Less commonly, meningovascular disease affects the anterior spinal artery and thereby causes spinal cord infarction.

The CSF abnormalities associated with meningovascular neurosyphilis are generally less severe than in those of acute meningitis, with lymphocytic pleocytosis between 10 to 100 cells/microL and protein concentration of 100 to 200 mg/dL. As with all of the early symptomatic forms of neurosyphilis, CSF-VDRL is usually but not always reactive.

Angiography (performed with computed tomographic, magnetic resonance, or conventional catheter techniques) may demonstrate focal segmental arterial narrowing, focal narrowing and dilatation, or total occlusion, similar to the findings seen in other infectious or noninfectious vasculitides. Neuroimaging shows one or more areas of infarction.

Late neurosyphilis — General paresis and tabes dorsalis are considered "tertiary" forms of neurosyphilis, while the early forms of neurosyphilis are not.

General paresis — General paresis (also known as general paralysis of the insane, paretic neurosyphilis, or dementia paralytica) is a progressive dementing illness. In the pre-penicillin era, general paresis resulted in death within an average of 2.5 years [4]. General paresis usually develops 10 to 25 years after infection, but it can occur as early as two years after infection. In the first half of the 20^th century, this form of neurosyphilis accounted for about 10 percent of all admissions to psychiatric hospitals.

In the early stage of disease, general paresis is associated with symptoms of forgetfulness and personality change. Most affected individuals experience progression of deficits in memory and judgment leading to severe dementia. Less often, patients may develop psychiatric symptoms such as depression, mania, or psychosis. In a study of 116 patients with general paresis in China, dementia, personality change, abnormal behavior, and emotional problems were the most common findings. On initial evaluation, neurosyphilis was not suspected in 36 percent of patients, which delayed the diagnosis for 1 to 24 months [25].

While the neurologic examination may be normal in some patients with general paresis, common abnormal findings include dysarthria, facial and limb hypotonia, intention tremors of the face, tongue, and hands, and reflex abnormalities. Pupillary abnormalities, including Argyll-Robertson pupils, may also be seen. However, pupillary findings are more typical of tabes dorsalis. (See 'Tabes dorsalis' below.)

CSF abnormalities are the rule in paretic neurosyphilis, with elevated lymphocytes of 25 to 75 cells/microL and a protein concentration in the range of 50 to 100 mg/dL. The CSF-VDRL is reactive in virtually all patients, although rare cases with a nonreactive CSF-VDRL have been reported [26]. Neuroimaging most commonly shows atrophy.

Tabes dorsalis — Tabes dorsalis (also called locomotor ataxia) is a disease of the posterior columns of the spinal cord and of the dorsal roots. It has the longest latent period between primary infection and onset of symptoms of all forms of neurosyphilis, with the interval averaging about 20 years, but sometimes as few as three years.

While tabes dorsalis was the most common form of neurosyphilis in the pre-antibiotic era, it is uncommon in the antibiotic era. (See 'Epidemiology' above.)

The most frequent symptoms of tabes dorsalis are sensory ataxia and lancinating pains. The latter are characterized by sudden, brief, severe stabs of pain that may affect the limbs, back, or face and that may last for minutes or days. Less common symptoms are paresthesia and gastric crises, characterized by recurrent attacks of severe epigastric pain, nausea, and vomiting. Bladder dysfunction with urinary retention and overflow incontinence may occur early in the course of disease.

Pupillary irregularities are among the most common signs in patients with tabes dorsalis, and the Argyll-Robertson pupil accounts for approximately one-half of these. An Argyll-Robertson pupil is small, does not respond to light, contracts normally to accommodation and convergence, dilates imperfectly to mydriatics, and does not dilate in response to painful stimuli. (See "Tonic pupil", section on 'Etiologic evaluation'.)

Other findings seen with tabes dorsalis include absent lower extremity reflexes, impaired vibratory and position sensation, and, less commonly, impaired touch, pain, and optic atrophy.

The CSF may be completely normal in tabes dorsalis, or may show a mild lymphocytic pleocytosis with 10 to 50 cells/microL and protein concentrations of 45 to 75 mg/dL. In this form of neurosyphilis, as many as one-quarter of the CSF-VDRL tests are nonreactive.

Atypical neurosyphilis — Some studies from the antibiotic era have reported "atypical" forms of neurosyphilis, a term that is used to describe neurosyphilis that does not fulfill the clinical criteria for one of the "classic" forms (eg, symptomatic meningitis, meningovascular syphilis, general paresis, and tabes dorsalis). As an example, a study from the early 1970s of 289 patients with reactive serum CSF fluorescent treponemal antibody absorption (FTA-ABS) tests reported that neurosyphilis was diagnosed in 241 (84 percent) based on the identification of reflex, sensory, or pupillary changes, or seizures [27]. However, similar cases of neurosyphilis were described in the 1940s [4]. Thus, it is likely that such "atypical" forms of neurosyphilis have always existed.

That said, several modern case reports have described patients with neurosyphilis that mimicked herpes encephalitis or autoimmune encephalitis [28,29]. These patients generally had acute onset of cognitive changes, although some patients had more slowly progressive presentations. Many had seizures. Brain MRI showed unilateral or bilateral medial temporal lobe lesions (image 2) characterized by high signal on T2 and fluid-attenuated inversion recovery (FLAIR) sequences, which resolved after treatment for neurosyphilis [28]. In addition to mesiotemporal lobe lesions, other reports have observed MRI signal abnormalities in parietal, temporo-occipital, and thalamic regions [29,30].

It is difficult to categorize these cases into one of the classic clinical descriptions of neurosyphilis described above (eg, symptomatic meningitis, meningovascular syphilis, general paresis, tabes dorsalis). While patients with a more slowly progressive course could be considered to have general paresis, those with acute onset and temporal lobe imaging abnormalities present a greater challenge for classification. The features of these cases are most similar to meningovascular syphilis, except that the temporal lobe abnormalities do not respect vascular territories. Thus, it is probably best to describe them as an overlap between meningeal and parenchymal disease, rather than attribute them to one of the classic clinical forms of neurosyphilis. The fact that seizures themselves may cause reversible temporal lobe abnormalities further complicates categorization of these cases [31].

DIAGNOSIS — In his 1944 text, Stokes wrote, "The frequency of neurosyphilis in general medical practice depends to a large extent on the thoroughness of the search for signs of neuraxis involvement and the frequency with which the spinal fluid examination is employed" [7]. This axiom remains true today. As an example, in a study from Seattle, WA, that systematically queried 68 patients with syphilis for symptoms of neurosyphilis, 8 percent had vision or hearing symptoms and approximately 4 percent had symptoms and objective confirmation of neurosyphilis [32]. The proportion with symptoms or neurosyphilis did not differ by HIV status. In another study of cerebrospinal fluid (CSF) abnormalities in individuals with syphilis, reactive CSF-Venereal Disease Research Laboratory (VDRL), diagnostic of neurosyphilis, was significantly more common in persons with HIV who had photophobia, vision loss, gait incoordination, or moderate or greater hearing loss [33]. No relationships between reactive CSF-VDRL and symptoms were identified in HIV-uninfected individuals.

Clinical suspicion and spinal fluid examination are keys to the diagnosis of neurosyphilis (algorithm 1 and algorithm 2).

●Lumbar puncture should be considered in the evaluation of a patient who presents with neurologic, otologic or ocular symptoms that could be caused by syphilis, but with unknown syphilis history.

●Lumbar puncture should be performed in the evaluation of a patient with known history of or current syphilis who presents with neurologic, otologic or ocular symptoms that could be caused by syphilis.

Current guidelines from the Centers for Disease Control and Prevention (CDC) do not recommend lumbar puncture in patients with syphilis who do not have neurologic, otologic, or ocular symptoms, stating "CSF examination has not been associated with improved clinical outcomes in the absence of neurologic signs and symptoms" [34]. This statement is true only because no outcomes research has been published. The issue of lumbar puncture in asymptomatic persons with HIV who have syphilis is addressed below.

Unknown syphilis history — In the setting of an unknown syphilis history, the first step in establishing the diagnosis of neurosyphilis is confirming that the patient has been infected with T. pallidum, as a patient cannot have neurosyphilis without first having syphilis.

Tests for syphilis include the following:

●Serum nontreponemal tests

•Venereal disease research laboratory (VDRL)

•Rapid plasma reagin (RPR)

●Serum treponemal tests

•Fluorescent treponemal antibody absorption (FTA-ABS)

•T. pallidum particle agglutination assay (TPPA)

•T. pallidum enzyme immunoassay (TP-EIA) and chemiluminescence immunoassay (CIA); note that TP-EIA and CIA should be confirmed with an alternative treponemal test that uses a different method

Confirmation is straightforward when serum nontreponemal tests and treponemal tests are reactive, as they virtually always are in early neurosyphilis. (See "Syphilis: Screening and diagnostic testing".)

Unfortunately, the nontreponemal tests (VDRL and RPR) may be nonreactive in late neurosyphilis, particularly in tabes dorsalis. When there is suspicion for late forms of neurosyphilis, serum treponemal tests should always be performed. These tests remain reactive for life in virtually all individuals regardless of previous treatment. Reactivity of these serum tests confirms that the patient has had syphilis at some time, and that he or she is at risk for neurosyphilis.

Patients with nonreactive serum treponemal tests do not merit further evaluation for late neurosyphilis. (See "Syphilis: Screening and diagnostic testing".)

Known syphilis — In patients with known syphilis, a lumbar puncture with cerebrospinal fluid (CSF) examination should be performed in certain clinical situations to evaluate the possibility of symptomatic or asymptomatic neurosyphilis (algorithm 1 and algorithm 2). The CDC recommendations are that CSF examination should be performed in any of the following situations [34]:

●Neurologic or ophthalmic signs or symptoms in any stage of syphilis.

●Evidence of active tertiary syphilis affecting other parts of the body.

●Treatment failure (including failure of serum nontreponemal tests to fall appropriately) in any stage of syphilis.

A few caveats should be noted:

●Patients who have ocular syphilis or otosyphilis, with or without CSF abnormalities, should receive standard neurosyphilis treatment. Ocular and otologic syphilis may not be accompanied by CSF abnormalities. The value of lumbar puncture is that if CSF abnormalities are identified, they can be followed to assess efficacy of treatment. (See 'Treatment' below.)

●Some experts recommend lumbar puncture for all patients with concomitant HIV infection and syphilis, regardless of stage, particularly those with serum RPR ≥1:32, peripheral blood CD4+ T cell count ≤350/microL, detectable plasma HIV RNA, or those not on antiretroviral therapy. These factors increase the odds of asymptomatic neurosyphilis three- to sixfold.  

Controversy exists regarding whether identification and treatment of asymptomatic neurosyphilis results in better patient outcomes. While data in the pre-antibiotic era suggested a benefit, no study has been performed in the modern era to address this question. We suggest lumbar puncture for persons with HIV and syphilis who have increased risk of asymptomatic neurosyphilis listed above, in agreement with European guidelines [35].

Spinal fluid examination — While a reactive CSF-VDRL establishes the diagnosis of neurosyphilis, a nonreactive test does not exclude the diagnosis [36]. Furthermore, the CSF-VDRL test may be falsely positive when the CSF is visibly blood-tinged and the serum nontreponemal test titer is high [37]. Reports of CSF-VDRL sensitivity and specificity vary depending on how neurosyphilis is defined, but the preponderance of data suggest that the test lacks sensitivity. In a monograph on neurosyphilis published during the 1940s when the disease was common, a nonreactive CSF Wassermann test, the predecessor of the CSF-VDRL, was documented in 8 to 19 percent of patients with early neurosyphilis, including meningitis and meningovasculitis, the forms that are most common today [4]. In a modern-era study of 149 patients with laboratory and clinically defined neurosyphilis, the diagnostic sensitivity of the CSF-VDRL ranged from 67 to 72 percent [38].

Because the CSF-VDRL is not available in some areas of the world, the CSF-RPR is sometimes used in its place. The CSF-RPR may be nonreactive in instances where the CSF-VDRL is reactive, making it a less useful test. The CDC syphilis guidelines do not address use of the CSF-RPR for neurosyphilis diagnosis [34]. The European guidelines state that the CSF-VDRL is preferred over the CSF-RPR [35].

In contrast to the CSF-VDRL, the CSF FTA-ABS test is sensitive but not specific, particularly in asymptomatic neurosyphilis. In the setting of lymphocytic CSF pleocytosis and a nonreactive CSF-VDRL, a nonreactive CSF FTA-ABS test excludes the diagnosis of asymptomatic neurosyphilis in most instances [39].

Elevation of the CSF protein concentration is consistent with neurosyphilis. However, it may be less specific than CSF pleocytosis [40,41].

●In patients with suspected neurosyphilis who do not have HIV infection and who have a nonreactive CSF-VDRL, a CSF lymphocyte count >5 cells/microL or a protein concentration >45 mg/dL is consistent with the diagnosis of neurosyphilis (algorithm 1).

●In patients who have HIV infection and syphilis, establishing a diagnosis of neurosyphilis is particularly difficult when the CSF-VDRL is nonreactive in the setting of a mild CSF pleocytosis. The difficulty arises because HIV itself causes mild CSF pleocytosis and mild elevation of the CSF protein concentration.

One study found that HIV-induced CSF pleocytosis was independently and significantly associated with three factors [42]:

●Lack of current antiretroviral use (OR 5.9, 95% CI 1.8-18.6)

●CD4 count >200/microL (OR 23.4, 95% CI 3.1-177.3)

●Detectable plasma HIV RNA (OR 3.3, 95% CI 1.1-9.4)

In persons with HIV who are taking antiretroviral agents, have CD4 counts ≤200 cells/microL, or have an undetectable plasma HIV RNA viral load, the likelihood of HIV-induced CSF pleocytosis is reduced by 70 to 96 percent [42]. Thus, CSF pleocytosis in such patients who have syphilis is more likely to be due to neurosyphilis than to HIV.

CSF concentration of the B cell chemokine CXCL13 may be useful in distinguishing between CSF pleocytosis due to syphilis and CSF pleocytosis due to HIV, but this test is not widely available [43,44].

TREATMENT — The treatment of neurosyphilis is discussed separately and reviewed here briefly. (See "Syphilis: Treatment and monitoring", section on 'Treatment of neuro/ocular/otic syphilis'.)

Antibiotic regimens

●Neurosyphillis - The standard regimens recommended by Centers for Disease Control and Prevention (CDC) guidelines for the treatment of neurosyphilis, including ocular and otosyphilis, are as follows (table 1) [34]:

•Aqueous crystalline penicillin G (18 to 24 million units per day, administered as 3 to 4 million units intravenous [IV] every four hours, or 18 to 24 million units daily as a continuous infusion) for 10 to 14 days, or

•Procaine penicillin G (2.4 million units intramuscular [IM] once daily) plus probenecid (500 mg orally four times a day), both for 10 to 14 days

These regimens can also be used following penicillin desensitization for patients who have a serious penicillin allergy [34]. Desensitization should be managed in consultation with a specialist (see "Penicillin allergy: Immediate reactions", section on 'Desensitization').

An alternative treatment for patients who have a mild penicillin allergy is ceftriaxone (2 g IV or IM daily) for 10 to 14 days, with careful observation for cross-reactivity [34,35,45]. High dose doxycycline (200 mg orally twice a day for 21 to 28 days) also may be an effective alternative treatment [46], but is not recommended by the CDC or European guidelines [34,35].

No controlled trials have evaluated the efficacy of the currently used forms of penicillin for neurosyphilis treatment. Thus, recommendations for treatment are based on clinical experience and on predicted penetration of penicillin into the cerebrospinal fluid (CSF) and central nervous system.

Neurological worsening after beginning treatment for neurosyphilis has been attributed to a Jarisch-Herxheimer reaction, and may be most common in patients treated for syphilitic dementia [47,48].

●Ocular syphilis and otosyphilis are treated with the same regimen as is used for neurosyphilis, regardless of the presence or absence of CSF abnormalities. Treatment with oral or topical glucocorticoids for ocular syphilis and oral glucocorticoids for otosyphilis is often used in conjunction with antibiotics.

Monitoring — Because T. pallidum cannot be routinely grown in the laboratory, there is no microbiological test of cure. Thus, the success of neurosyphilis therapy is judged by resolution or stabilization of clinical abnormalities and by normalization of CSF abnormalities.

Neurologic examination and lumbar puncture should be performed at three to six months after treatment and every six months thereafter until the CSF white blood cell count is normal and the CSF-Venereal Disease Research Laboratory (VDRL) is nonreactive. We suggest that failure of the CSF white blood cell count to decrease six months after therapy or failure of CSF-VDRL to decline fourfold (or to nonreactive if the initial titer is <1:2) one year after therapy are indications for re-treatment. The CSF white blood cell count should decline by six months after successful treatment, and all CSF abnormalities should resolve by two years after treatment. Failure to meet these criteria should prompt retreatment. Retreatment is also indicated if any follow-up CSF sample shows an increase in CSF white blood cell count, or a fourfold increase in CSF-VDRL titer. CSF abnormalities may normalize more slowly in patients who have HIV infection compared with those who are not infected with HIV [40].

Normalization of serum rapid plasma reagin (RPR) titer may predict successful neurosyphilis treatment. Supporting evidence comes from a longitudinal study of 110 patients with neurosyphilis, most with HIV coinfection, who were treated for neurosyphilis [41]. At 4, 7, and 13 months after treatment, normalization of the serum RPR test, defined as a fourfold or greater decline in RPR titer or reversion to nonreactive, predicted normalization of CSF white blood cell count and CSF-VDRL reactivity in >90 percent of patients. However, the predictive value of serum RPR normalization was lower for persons with HIV who were not taking antiretroviral agents

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sexually transmitted infections".)

SUMMARY AND RECOMMENDATIONS

Epidemiology and clinical manifestations

●Neurosyphilis refers to infection of the central nervous system (CNS) by Treponema pallidum. This disorder begins with invasion of the cerebrospinal fluid (CSF), a process that probably occurs shortly after acquisition of T. pallidum infection. (See 'Pathogenesis' above.)

●Early in the course of syphilis, the most common forms of neurosyphilis involve the CSF, meninges, and vasculature (asymptomatic meningitis, symptomatic meningitis, and meningovascular disease). Later in disease, the most common forms involve the brain and spinal cord parenchyma (general paresis and tabes dorsalis) (figure 1). In the current era, early neurosyphilis is more common than late neurosyphilis, and is most frequently seen in people with HIV. (See 'Epidemiology' above and 'Clinical manifestations' above.)

●Asymptomatic neurosyphilis lacks symptoms or signs of CNS disease, although there may be concomitant evidence of primary or secondary syphilis. Treatment of asymptomatic neurosyphilis prevents progression to the symptomatic forms. (See 'Asymptomatic neurosyphilis' above.)

●Symptomatic syphilitic meningitis is typically manifested by headache, confusion, nausea and vomiting, and stiff neck. Visual acuity may be impaired if there is concomitant uveitis, vitreitis, retinitis, or optic neuropathy. Other signs include cranial neuropathies. (See 'Symptomatic meningitis' above.)

●Syphilitic meningitis can cause an infectious arteritis that may affect any vessel in the subarachnoid space and result in thrombosis, ischemia, and stroke involving the brain or spinal cord. (See 'Meningovascular syphilis' above.)

●The late forms of neurosyphilis (general paresis and tabes dorsalis) are the "tertiary" forms, while the early forms of neurosyphilis are not.

●General paresis is a progressive dementing illness that usually develops 10 to 25 years after infection. Symptoms of forgetfulness and personality change predominate early on, but progression leads to severe dementia. Common signs include dysarthria, facial and limb hypotonia, intention tremors of the face, tongue, and hands, and reflex abnormalities. (See 'General paresis' above.)

●Tabes dorsalis, uncommon in the modern era, is a disease of the posterior columns of the spinal cord and of the dorsal roots. The latent period between primary infection and onset of symptoms averages about 20 years. The most frequent manifestations include sensory ataxia, lancinating pains, and pupillary irregularities. (See 'Tabes dorsalis' above.)

Diagnosis summary

●Clinical suspicion and spinal fluid examination are keys to the diagnosis of neurosyphilis. In the setting of unknown syphilis history, the first step is confirming current or previous infection with T. pallidum (algorithm 1 and algorithm 2). Serum nontreponemal tests (Venereal Disease Research Laboratory [VDRL] and rapid plasma reagin [RPR]) and serum treponemal tests (fluorescent treponemal antibody absorption [FTA-ABS], T. pallidum particle agglutination assay [TPPA], or T. pallidum enzyme immunoassay [TP-EIA] or chemiluminescence immunoassay [CIA]) are virtually always reactive in early neurosyphilis. However, the nontreponemal tests may be nonreactive in late neurosyphilis, while the treponemal tests remain reactive. Therefore, serum treponemal tests should always be performed when there is suspicion for late forms of neurosyphilis. Note that a positive TP-EIA or CIA should be confirmed with an alternative treponemal test that uses a different method. (See 'Diagnosis' above.)

●In a patient with known syphilis, we recommend lumbar puncture with CSF examination for those who have neurologic, otologic, or ocular signs or symptoms, active tertiary syphilis affecting other parts of the body, or treatment failure in any stage of syphilis. In addition, we suggest lumbar puncture for persons with HIV and syphilis who are neurologically asymptomatic but at increased risk of neurosyphilis as determined by a serum RPR titer ≥1:32, a peripheral blood CD4+ T cell count ≤350/microL, the presence of detectable plasma HIV RNA, or nontreatment with antiretroviral therapy. (See 'Known syphilis' above.)

●The CSF-VDRL is specific but not sensitive for the diagnosis of neurosyphilis. In contrast, the CSF FTA-ABS test is sensitive but not specific, particularly in asymptomatic neurosyphilis. Establishing a diagnosis of neurosyphilis is difficult when the CSF-VDRL is nonreactive in patients with concurrent HIV infection and syphilis who have a mild CSF pleocytosis, because HIV itself can cause a mild CSF pleocytosis and elevation of the CSF protein concentration. (See 'Spinal fluid examination' above.)

Treatment summary

●For patients with neurosyphilis who are not penicillin-allergic, we recommend treatment with penicillin (Grade 1B). The standard regimens recommended by the Centers for Disease Control and Prevention (CDC) are listed above and shown in the table (table 1). Patients who have a serious penicillin allergy must first undergo desensitization prior to penicillin therapy. For patients who have a mild penicillin allergy, we suggest ceftriaxone (2 g intravenous daily) for 10 to 14 days with careful observation for cross-reactivity (Grade 2C). For patients with serious penicillin allergies who refuse desensitization or those who cannot be treated with penicillin and ceftriaxone, an alternative is oral doxycycline (200 mg twice daily) for 21 to 28 days. (See 'Treatment' above.)

●Lumbar puncture should be performed three to six months after treatment and every six months thereafter until the CSF white blood cell count is normal and the CSF-VDRL is nonreactive. We suggest retreatment if the CSF white blood cell count does not decrease six months after therapy or the CSF-VDRL does not decline fourfold (or to nonreactive if the initial titer is <1:2) one year after therapy. In addition, we suggest retreatment if there is an increase in the CSF white blood cell count, or a fourfold increase in CSF-VDRL titer, in any follow-up CSF sample. (See 'Monitoring' above.)

●In immunocompetent patients or patients with HIV who are taking antiretroviral agents and in whom a follow-up lumbar puncture cannot be performed, normalization of the serum RPR titer can be used as a surrogate for success of neurosyphilis treatment. (See 'Monitoring' above.)