Red Pimple-like Rashes on Body


My friend who was recently diagnosed HIV+ was also found to have syphilis. His CD4 and viral load status are healthy so he is not on any medication now. However, he went on a penicillin treatment to cure his syphilis. After the penicillin treatment was completed, he began to develop a fever and red pimple-like rashes (some itchiness) formed on his body. I read somewhere that syphilis will somehow suppress the HIV virus. Hence, we relate the fever and rash outbreak to the syphilis being cured. Are we right to say that? And is this normal and is it ok to leave the rashes untreated, assuming that it will subside after a while? Or do we need to undertake some treatment for this? He's currently under some general medicine to treat his fever.



The interaction between HIV and syphilis is complex. (See below.)

What you are describing may be a Jarish-Herxheimer reaction. This can occur when secondary syphilis is treated with penicillin and appears to be more common in HIV-positives than neggies. The reaction is characterized by fever, fatigue and transient worsening of any mucocutaneous (skin and mucous membrane) symptoms. It usually subsides within a day or so.

Your friend should review the information below and then talk to his HIV specialist.

Good luck!

Dr. Bob

Treatment for syphilis (SYPHILIS) Apr 8, 2008

Dear Dr Bob,

I was recently diagnosed with HIV on 22nd Feb 08.

After a few blood tests, my HIV specialist informed me that I have also contracted syphilis.

I was told that I need to redo a CD4 count again to confirm further treatment for syphilis.

If my CD4 count is normal, I will have to undergo a lumbar puncture procedure to confirm if the syphilis virus has spread to my brain. If the virus has spread to my brain, then I will need to be admitted to hospital for 2 weeks to be put on the drip to administer proper medicine to kill the virus. If the virus has not spread to the brain, then basically I just need to be given 3 jabs; 1 jab per week which is sufficient to kill the virus.

If my CD4 count is below normal, then I will have to go for a brain scan and my HIV specialist will only be able to determine the next course of action to treat my syphilis infection on the outcome of the scan report.

I am really lost at this point of time as I am not sure if this is the right course of action to be administered to patients suffering from syphilis coupled with HIV infection. Is there a simpler way to treat syphilis without going for lumbar puncture procedure? I am also worried if I am to be admitted to hospital for 2 weeks; my company will come to know of my HIV infection and I may very well lose my job. Also I am extremely worried about all the costs involved as medical treatment in Singapore is expensive.

I hope you can help me out and give me proper guidance what other alternatives do I have?

Thanks a lot.


Response from Dr. Frascino

Hello JC,

The interaction of HIV disease and syphilis is complex, to say the least. Without a much more detailed history, review of your laboratory studies and the benefit of a physical examination, it will be difficult for me to provide very specific recommendations regarding additional evaluation and optimal treatment for your coinfection. I will make some general comments and also provide a detailed reference below.

Regarding the need for a lumbar puncture, evaluation of cerebral spinal fluid for evidence of neurosyphilis is recommended for anyone with neurologic, ophthalmic or otologic signs or symptoms; all patients with late latent syphilis lasting more than one year or syphilis of unknown duration. If you fall into any one of those categories, a lumbar puncture is warranted no matter what your CD4 count is. Brain scans are not routinely recommended for evaluation of syphilis. Late latent syphilis is diagnosed by a positive blood test without physical signs or symptoms of either the primary or secondary stages of syphilis. (As I mentioned this can be quite a confusing and complex topic!)

Regarding treatment, intravenous penicillin G for 10-14 days is recommended for neurosyphilis. Primary, secondary, early latent and late latent syphilis in HIV-infected patients is treated with intramuscular injections of benzathine penicillin G.

A more detailed discussion of syphilis/HIV coinfection is reprinted below. This may be more information than you wanted. If you are uncertain of your current physician's recommendations, I would suggest you get a second opinion. You should stress the need for confidentiality with your physicians in Singapore to protect your privacy and employment. That an illness would be cause for job termination is reprehensible.

Good luck.

Dr. Bob

Syphilis and HIV

HIV InSite Knowledge Base Chapter June 2006

Introduction and Overview

This chapter discusses the epidemiology, clinical manifestations, and treatment of syphilis in individuals with HIV. The interaction of syphilis and HIV infection is complex and remains the subject of ongoing research.(1-6) Epidemiologic studies demonstrate that sexually transmitted diseases (STDs) including syphilis, and particularly genital ulcers associated with primary syphilis, are associated with an increased risk of HIV acquisition.(7-12) Although case reports have suggested that coexisting HIV infection may alter the natural history of syphilis, only a few such effects have been demonstrated in large observational studies,(13-19) notably presentation with multiple or deeper chancres and overlap of primary- and secondary-stage features of syphilis in coinfected patients.(20,21) Initial serologic responses to early syphilis were shown to be generally equivalent in HIV-negative and -positive individuals.(20) Like many acute infections in the HIV-infected patient, early syphilis may decrease CD4+ T-cell counts (CD4 cell counts) and increase HIV RNA in plasma and semen, although evidence for these effects has been mixed.(22-24) Reports of false-positive and false-negative results on serologic tests for syphilis in HIV-infected persons raise questions regarding the specificity and sensitivity of serologic diagnoses in such patients.(25-32) Obtaining biopsy specimens from suspicious lesions and using special stains for spirochetes should be considered in patients who are seronegative for syphilis. Cerebrospinal fluid (CSF) abnormalities in early syphilis and persistence of nontreponemal antibodies after treatment have both been observed in HIV infection, raising concerns about a possible increased risk of clinically important neurologic disease.(25-27,29-31,33-44)

Routine, periodic screening for STDs in HIV-infected patients is strongly recommended, including serologic testing for syphilis.(45,46) Although data from prospective, controlled treatment studies are limited, this chapter presents recommendations for managing HIV-infected patients with syphilis based on the best available data.(47) Data suggest that current recommendations for treating early syphilis of less than 1 year in duration are adequate for most patients, whether or not they have HIV infection. Long-acting penicillin G is the antibiotic of choice. The only oral alternatives recommended by the U.S. Centers for Disease Control and Prevention (CDC)--doxycycline and tetracycline--should be reserved for those few who are truly allergic to penicillin, and such agents are never indicated for the treatment of neurosyphilis. One agent, azithromycin, has shown promise for incubating and early syphilis;(48,49) however, given the recent identification of resistance, it cannot be recommended for general use in the United States.(50,51) Follow-up serologic testing at recommended intervals is essential for assessing treatment adequacy;(52) these assessments are often complicated by the potential for reexposure in sexually active persons. Evaluation of the CSF for evidence of neurosyphilis is recommended for any patient with neurologic, ophthalmic, or otologic signs or symptoms; all patients with treatment failure; and HIV-infected patients with late latent syphilis lasting more than 1 year or with syphilis of unknown duration. Options to consider in some HIV-infected patients, although of unproven benefit, include increasing the duration of penicillin treatment in patients with early syphilis, evaluating the CSF for evidence of neurosyphilis earlier in the course of syphilis infection or after treatment, and evaluating the CSF in patients with lower CD4 cell counts or a nontreponemal serologic test titer greater than 1:32.


Epidemiologic studies demonstrate that history of an STD, including syphilis, is associated with an increased risk of HIV disease among both gay men (7,9) and heterosexuals (8) because sexual behaviors that increase the risk of acquiring STDs also increase the risk of acquiring HIV. Furthermore, genital ulcerations and inflammation caused by STDs are implicated as cofactors for acquiring or transmitting HIV infection.(9-11) Recent data suggest that when they already have other STDs, individuals are 3 to 5 times more likely to acquire HIV if exposed to the virus through sexual contact.(10)

In U.S. areas reporting high rates of syphilis infection in the early 1980s, the percentage of early syphilis cases occurring among gay and bisexual men decreased from 50% to 70% in the late 1970s to 5% to 15% in 1990.(53) These data were thought to reflect changes in sexual practices that reduced the risk of HIV infection among gay and other men who have sex with men (MSM) and suggested that education efforts encouraging safer-sex practices were effective. During the 1990s, the geographic distribution of heterosexual HIV transmission closely paralleled that of syphilis. Most of the health districts with the highest rates of syphilis in 1995 were concentrated in the southern United States, where the HIV prevalence among childbearing women also was highest. Therefore, increases in the incidence of STDs in any population may presage future increases in HIV, illustrating the importance of STD diagnosis and treatment as an HIV preventive strategy.

Despite significant decreases in cases of syphilis in the United States in the 1990s, as well as the CDC's initiative to eliminate the disease domestically by 2005, syphilis incidence has risen dramatically since 2000, particularly among MSM.(3,54-60) (See the CDC's Syphilis Surveillance Project Annual Report for a detailed epidemiologic update.) In 2004, 7,980 cases of primary and secondary syphilis were reported nationally, representing an 11.2% increase over 2003 and a 33% increase over the previous decade's nadir of 5,979 cases, reported in 2000.(61) A number of states have experienced particularly large recent increases in syphilis, concentrated among MSM. In 2005 in California, for instance, 1,489 cases of primary and secondary syphilis were reported, representing an increase of more than 700% from 1999. Of the 2005 cases, 79% were MSM and of those, 61% were HIV positive by self-report, and a majority (60%) reported multiple anonymous sex partners.(62) Recently, the Syphilis Elimination Effort has been launched as a national initiative to bring together health care providers, policy makers, community leaders, and state and local public health agencies to reduce syphilis rates in the United States.

Suggested causes of this resurgence in syphilis among MSM include increases in the number of anonymous sex partners, decreases in condom use, use of the Internet for meeting sex partners, and more widespread use of methamphetamine and proerectile agents, among other drugs.(4,63-67) Oral sex has been implicated in 14% of syphilis infections in a large U.S. city.(68) Moreover, the indirect contribution of antiretroviral treatment (ART) to higher syphilis transmission rates among those infected with HIV has been well described.(69) These epidemiologic trends underscore the importance of clinicians' awareness of key behavioral risk factors, their consideration of syphilis in the differential diagnosis of similarly presenting disorders, and their familiarity with current recommendations regarding risk assessment, screening, and clinical management of syphilis.

Limited epidemiologic data suggest that neurosyphilis remains an uncommon infection in HIV-infected patients. A retrospective surveillance investigation of neurosyphilis trends in San Francisco from 1985 through 1992 found no evidence of increasing rates of neurosyphilis during the study interval.(70) Reports of symptomatic neurologic involvement totaled 2 to 5 cases per year. No cases of early neurosyphilis were identified in HIV-uninfected patients. More recent surveillance has confirmed a consistently low incidence of neurosyphilis in San Francisco, found in 2.7% (33/1,232) of all syphilis cases (all stages) in 2002 to 2003.(71)


The immune system plays an important role in protecting against syphilis.(72,73) Impairment of both cell-mediated and humoral immunity by HIV (74) may limit the host's defenses against Treponema pallidum, thereby altering the clinical manifestations or natural course of syphilis infection. In animal models, selective impairment of cell-mediated immunity shortens the incubation time of syphilis, increases the number and spreading of lesions, and slows healing time.(75) Molecular pathogenetic mechanisms to explain the role of the spirochete in facilitating HIV transmission may include up-regulation of gene expression, such as that of the CCR5 co-receptor used in HIV entry.(76,77) HIV-induced meningeal inflammation may facilitate penetration of spirochetes into the central nervous system (CNS) and thus contribute to the development of symptomatic neurosyphilis.

Clinical Manifestations and Course

Case reports suggest that the unusual clinical manifestations of syphilis may be more common and the course more rapid in patients with HIV infection. Although prospective studies largely have not corroborated these observations, such anecdotal reports have driven the hypotheses that among patients coinfected with HIV and T pallidum, cutaneous lesions may be more severe, symptomatic neurosyphilis may be more likely to develop, the latency period before the development of meningovascular syphilis may be shorter, and the efficacy of standard therapy for early syphilis may be reduced.

Primary and Secondary Syphilis

The conventional staging of syphilis is unaltered by HIV coinfection. Symptoms may appear within several days of exposure (mean, 21 days; range, 14-90 days).(4,78) In primary syphilis, 1 or more ulcers (chancres) develop, presumably at the site of exposure, usually on the genitals or anus. Patients may be unaware of these typically painless lesions, especially if located inside the vagina or anus. Regional, shotty, nontender lymphadenopathy may develop. If the ulcer or ulcers become infected secondarily, then painful lesions, adenopathy, or both may be observed. The chancre(s) then resolves, with or without treatment. Most primary lesions resolve by the time secondary syphilis presents; however, the primary and secondary stages may overlap.

In the secondary stage, symptoms typically include generalized or localized skin eruptions and mucosal lesions. Lesions may be mild or florid and usually include a wide morphologic variety of rashes--ranging from macular, papular, pustular, papulosquamous, to annular--that may cover the palms and soles. Constitutional symptoms (eg, sore throat, low-grade fever, malaise, myalgias, arthralgias, generalized lymphadenopathy) also may be present. Less common symptoms and signs include mucous patches in the oropharynx, condylomata lata (moist, wartlike papules occurring mostly in genital or perianal areas), and patchy alopecia. The rash of secondary syphilis can mimic many dermatologic conditions, such as tinea versicolor, pityriasis rosea, scabies, fixed drug eruptions, and erythema multiforme; in HIV-infected individuals taking ART, it has been misdiagnosed as an antiretroviral drug reaction.

In general, "infectious syphilis" refers to the primary and secondary stages because syphilitic chancres, mucous patches, and condylomata lata are highly infectious lesions. Dry rashes, by exception, are typically noninfectious. Like primary lesions, secondary symptoms and signs will resolve without therapy; such resolution marks the beginning of the latent period of disease.

Most HIV-infected patients with T pallidum infection present with typical dermatologic clinical features of primary and secondary disease, such as chancres and diffuse maculopapular rashes.(20,34,40,79,80) In a study among patients seen in an STD clinic, however, patients with HIV infection were more likely than HIV-negative individuals to present with signs and symptoms of secondary syphilis and were more likely to have chancres still present at the time of secondary syphilis diagnosis.(80) In a multicenter study of STD clinic patients with early syphilis, HIV-infected patients were more likely to present with multiple chancres, but the size of the chancres, characteristics of the skin rash, and duration of the chancres or rash before presentation did not differ according to HIV status.(20) Atypical chancres have been reported,(81) including lesions appearing as fissures or abrasions. Two reports also described gummatous penile ulceration.(82,83) Unusual rashes include papular or nodular eruptions,(27,84) nodular or ulcerative lesions with necrotic centers (ie, lues maligna),(18,19,27,28,85-90) and keratoderma.(91) These skin lesions have been characterized as more aggressive forms of secondary syphilis in HIV-infected persons, yet the same dermatologic presentations are found in HIV-uninfected persons.(92) These isolated findings do not reveal the frequencies of these uncommon cutaneous manifestations, and therefore, whether the clinical course and clinical spectrum of syphilis differ in HIV-infected and HIV-uninfected populations remains unclear. In general, syphilis should be included in the differential diagnosis of mucocutaneous abnormalities in the HIV-infected patient, and diagnostic evaluation and empiric treatment should be applied as discussed below.

Latent and Tertiary Syphilis

In latent syphilis, serologic evidence of infection is found despite absence of symptoms and signs of the primary and secondary stages. Relapses of secondary syphilis symptoms and signs may occur early in the latent stage. Early latent syphilis is defined by CDC as infection less than 12 months in duration, evidenced in the previous year by a negative serologic test, symptoms or signs of primary or secondary syphilis, or contact with a sex partner with early-stage syphilis.(47) If none of these criteria suggest that infection was acquired in the preceding 12 months, then the duration of infection is unclear and the case is referred to as syphilis of unknown duration. In this case, even if the patient had a negative or 4-fold lower nontreponemal serologic test result more than 12 months ago, it is usually impossible to determine the time of acquisition since the previous test, assuming that risk behaviors occurred throughout this time period. When the precise time of acquisition is in doubt because the CDC's early latent criteria are not met, these patients should be treated conservatively as if the infection were present for more than 1 year. These cases require CSF evaluation and treatment with benzathine penicillin G, 7.2 million units divided in 3 weekly doses (ie, 2.4 million units per dose), assuming no evidence of neurosyphilis was found.

Tertiary syphilis refers to disease presenting with late manifestations, encompassing c ardiovascular features such as aortitis with aneurysm formation, late neurologic sequelae (general paresis, tabes dorsalis), and formation of gummas (indolent but potentially destructive granulomatous lesions that may occur in any organ but chiefly involve skin, bone, and liver). Of note, neurosyphilis is not a stage, but rather a site of infection, where symptoms may manifest either earlier or later in the course of infection.


Neurosyphilis refers to a site of infection involving the neurologic system. As such, neurosyphilis may occur at any stage of syphilis. Neurosyphilis in patients with HIV infection is well described.(15,70,93-97) Reports of neurosyphilis in HIV-infected persons are similar to cases reported before the HIV epidemic.(98-100) The precise extent and significance of neurologic involvement in HIV-infected patients with syphilis, reflected by either laboratory or clinical criteria, remain incompletely characterized. Furthermore, the alteration of host immunosuppression by ART in recent years has further complicated such characterization.

The clinical spectrum of neurosyphilis and the typical interval between primary infection and neurologic symptoms were described early in the HIV epidemic.(100) Approximately 35% to 40% of persons with secondary syphilis have asymptomatic CNS involvement, as demonstrated by any of the following on CSF examination: an abnormal cell count, protein level, or glucose level; or demonstrated reactivity to Venereal Disease Research Laboratory (VDRL) antibody test. Acute syphilitic meningitis usually occurs within the first 2 years of infection; 10% of cases are diagnosed at the time of the secondary rash. Patients present with headache, meningeal irritation, and cranial nerve abnormalities, typically involving cranial nerves at the base of the brain (especially II, VI, VII, and VIII).

Meningovascular syphilis occurs a few months to 10 years (average, 7 years) after the primary infection. Unlike the sudden onset of thrombotic or embolic stroke syndromes, meningovascular syphilis is associated with prodromal symptoms lasting weeks to months before focal deficits of a vascular syndrome are identifiable. Prodromal symptoms include unilateral numbness, paresthesias, extremity weakness, headache, vertigo, insomnia, and psychiatric abnormalities such as personality changes. The focal deficits initially are intermittent or progress slowly over a few days. By contrast, general paresis and tabes dorsalis are parenchymatous forms of neurosyphilis that occur, in general, 10 to 30 years after primary infection. General paresis causes symptoms similar to those of any dementia and of many psychiatric illnesses such as depression and paranoia. Tabes dorsalis is associated with a triad of symptoms (lightning pains, dysuria, and ataxia) and a triad of signs (the Argyll Robertson pupil [accommodation in the absence of light reaction], areflexia, and loss of proprioceptive sense).

Most cases of symptomatic neurosyphilis reported among HIV-infected persons are early forms of neurosyphilis; namely, acute syphilitic meningitis and meningovascular neurosyphilis.(70) A case report of medical medullary syndrome with meningovascular syphilis has been described in an HIV-infected man.(96) Tabes dorsalis has been reported in 1 HIV-infected man who had been exposed 7 years earlier (94) and in another man treated for primary syphilis 7 years earlier.(101) Other reports have described cases of syphilitic meningomyelitis with spastic paraparesis (102,103) and syphilitic polyradiculopathy with progressive leg pain and weakness.(104) Status epilepticus,(105) presenting with or without cerebral parenchymal lesions attributed to syphilis,(106-108) has been described.

Data suggest that asymptomatic CNS involvement at the time of early syphilis infection is common, but not more common in HIV-infected persons than in uninfected persons.(18,20,33,35) One study (35) found viable T pallidum in the CSF of 12 of 40 (30%) patients with primary or secondary syphilis and no neurologic symptoms, including 3 of 10 (30%) HIV-infected patients, and another report (20) identified T pallidum in the CSF in 32 of 131 (24%) patients with early syphilis, including 11 of 43 (26%) HIV-infected patients. Whether neurologic complications of syphilis occur more frequently and earlier in HIV-infected patients has not been evaluated definitively. Most cases of early neurosyphilis in HIV-infected individuals, however, are currently identified at the initial syphilis diagnosis,(70) and, in at least 1 series, among those with CD4 counts less than 200 cells/L.(109) Therefore, most experts believe that HIV-infected patients with early syphilis have an increased risk of neurologic complications, but the magnitude of any such risk is uncertain. Careful neurologic evaluation in any patient with syphilis or HIV is of paramount importance. If such evaluation reveals abnormalities, neurosyphilis should be included in the differential diagnosis.

Ocular and Otologic Neurosyphilis

Ocular and otologic manifestations of neurosyphilis in HIV-infected persons have been reported widely.(110-112) The most common ocular findings in patients with concurrent HIV infection are uveitis,(113-116) chorioretinitis,(117) and retrobulbar neuritis.(118) Reports have also described retinitis or neuroretinitis,(119,120) scleritis,(121,122) and papillitis, vitreitis, and optic perineuritis.(110,112) The most common presenting symptoms are decreased vision and eye pain. In addition to the abnormalities of the optic cranial nerve II and the ocular motor nerves III and VI that can occur with acute syphilitic meningitis, these other ocular manifestations of syphilis have commonly been associated with the secondary stage of syphilis infection and CNS involvement. One case report described a gumma involving the optic nerve.(123)

Otologic syphilis is one of the few forms of sensorineural hearing loss that can be treated effectively if detected early in the course, based on clinical suspicion followed by CSF evaluation and treatment for neurosyphilis. Although the incidence of otologic symptoms in patients with HIV infection is apparently low, 1 report described 5 cases of otosyphilis in persons with coexisting HIV infection.(111) Otologic findings in these patients included progressive hearing loss, tinnitus, imbalance, and a sensation of ear fullness. Three patients had been treated for primary syphilis 2 to 5 years before the onset of symptoms. Only 1 of these 5 patients had acute syphilitic meningitis and an abnormal CSF profile coincident with the diagnosis of otosyphilis. One case report described otosyphilis involving an internal auditory-canal gumma.(124)

These clinical manifestations of ocular and otologic syphilis have also been observed among persons without HIV infection. It is unknown whether ocular and otologic findings occur more frequently in HIV-infected persons. Careful ophthalmologic and otologic examinations of symptomatic HIV-infected persons and inclusion of syphilis in the differential diagnosis remain essential.

Other Clinical Manifestations of Syphilis

Syphilis is a systemic infection. Hematogenous dissemination of the organism occurs early in the course of infection, and disease ultimately may involve any organ system. Symptomatic involvement of organs other than the skin, mucous membranes, and CNS is rare, presumably because the treponeme replicates at relatively low temperatures. Nonetheless, case reports of HIV-infected patients with pneumonitis,(125) hepatitis,(126-128) osteitis,(129,130) and gastritis (131,132) due to syphilis have been published. Gummatous syphilis, a tertiary form of syphilis, also has been described in HIV-infected persons.(133-135)

Diagnosis of Syphilis in HIV-Infected Patients

The diagnosis of syphilis may be more complicated in HIV-infected patients because of false-negative and false-positive serologic results for T pallidum and atypical clinical presentations in the presence of HIV infection. However, serologic tests appear to be accurate and reliable for the diagnosis of syphilis and the evaluation of treatment response in most HIV-infected patients. The clinician should seek confirmatory evidence for the diagnosis from any available source, including the patient's history, clinical findings, direct examination of lesion material for spirochetes, and serologic tests for syphilis. The importance of a careful clinical examination of HIV-infected patients at risk for syphilis cannot be overstated. Painless primary syphilis lesions, especially involving the rectum and vagina, often go unnoticed, and secondary syphilis rashes, other mucocutaneous lesions, and constitutional symptoms may be subtle or misleading and resolve without treatment. CNS disease can occur during any stage of syphilis. Clinical evidence of neurologic involvement warrants CSF examination.

If at all feasible, the clinician should perform dark-field examination or direct fluorescent antibody (DFA) staining of exudate from suspicious lesions of primary syphilis. For patients with suspicious lesions but negative serologic results, positive findings on dark-field examination or DFA stain can be diagnostic. Clinicians should also use dark-field examination or DFA staining of selected secondary lesions, such as condylomata, to establish the diagnosis of secondary syphilis. However, it is important to confirm by DFA that the treponema seen in dark-field-positive oral lesions are T pallidum, because nontreponemal spirochetes occur in the upper gastrointestinal tract, especially the mouth. Although a multiplex polymerase chain reaction (PCR) test for genital ulcerative lesions has been developed, with excellent sensitivity (approximately 95%) for syphilis as well as herpes simplex virus and chancroid, there are no current plans to market this test in the United States.(136)

Serologic tests for syphilis are the cornerstone of diagnosing untreated syphilis infection, independent of HIV status. Nontreponemal assays, such as the rapid plasma reagin (RPR) card or VDRL test, use cardiolipin-, lecithin-, and cholesterol-containing antigen to measure antilipoidal antibodies and are often used initially to diagnose syphilis. Treponemal assays (Treponema pallidum particle agglutination [TP-PA] and fluorescent treponemal antibody absorbent [FTA-Abs]) are used to confirm the results of positive nontreponemal tests in a 2-step, reflex process. The sensitivity of nontreponemal and treponemal tests for syphilis increases with duration of infection, and ranges from approximately 75% in the primary stage to virtually 100% in the secondary stage.(137) In general, the sensitivity of treponemal tests continues to approximate 100% in late syphilis, in contrast to nontreponemal tests, which are more practical and cost-effective for initial screening but have diminished sensitivity in late syphilis. Despite the higher sensitivity of treponemal tests, they have not been recommended for initial screening in the United States, primarily because of cost.

Because the sensitivity of nontreponemal tests is lower that that of treponemal tests in the primary stage, a negative nontreponemal test in an HIV-infected individual with a genital lesion cannot exclude primary syphilis. It may be useful to consider both a nontreponemal and a (nonreflexed) treponemal test as a diagnostic strategy in newly infected persons with suspicious lesions. For instance, in a recent evaluation of 39 dark-field-confirmed syphilis cases, VDRL was reactive in 30 (77%), whereas TP-PA was positive in 37 (95%). Using a strategy of initially obtaining both VDRL and TP-PA would have identified 100% of primary syphilis cases, compared with only 72% of cases using a conventional reflex strategy.(138)

Unusual serologic responses have been reported in HIV-infected persons with syphilis. Most reports involved higher than expected serologic titers, but false-negative serologic results and delayed appearance of seroreactivity have also been reported, albeit rarely. Nevertheless, both treponemal and nontreponemal serologic tests for syphilis are accurate in most patients with syphilis and HIV coinfection.(47) Clinicians should obtain serum samples from any patient in whom the diagnosis of syphilis is suspected. Clinicians should also screen all sexually active patients with HIV infection for asymptomatic syphilis infection on a routine, periodic basis, generally every 6 months, if risk warrants.(45) Reports of nontreponemal antibody test results should be quantitative and describe the lowest dilution (ie, the titer) at which the test result is reactive. In general, nontreponemal test titers may be higher among HIV-positive patients than among HIV-negative persons.(29) In 1 study, the median RPR titer among HIV-infected patients with early syphilis was 1:128, in comparison with a median titer of 1:64 among HIV-negative patients.(20)

The specificity of the nontreponemal serologic tests for syphilis may be compromised in HIV-infected persons.(25,29,30,139) Very high VDRL or RPR titers of greater than 1:64 have been reported in HIV-infected patients without syphilis.(140) These nontreponemal tests detect antibodies directed against a cardiolipin-lecithin antigen. In patients with immunoglobulin abnormalities, the RPR or VDRL test result may be falsely positive. Many persons with HIV infection have both anticardiolipin-lecithin antibodies and polyclonal gammopathy. A positive RPR or VDRL test result may be biologically false positive if the confirmatory test is negative, and thus may not represent active syphilis infection. However, in patients epidemiologically at risk for syphilis infection, newly reactive nontreponemal and treponemal serologic syphilis test results usually indicate active disease. The serofast state is not uncommon among HIV-infected persons and refers to the persistence of a reactive nontreponemal syphilis test, usually 1:16 or less, with variation no greater than 1 to 2 dilutions over time, despite adequate therapy and no evidence of reexposure. However, reinfection may be difficult to rule out in some patients, and reactivation or relapse of a previously treated infection is also possible in a person with HIV infection.

A negative RPR or VDRL test result may not rule out syphilis in patients with HIV infection. Although the sensitivity of these serologic tests in diagnosing secondary syphilis is generally very high, rare case reports have described seronegative secondary syphilis in patients with HIV infection,(27,28,31,139,141) suggesting that some patients may fail to develop normal antibody responses to T pallidum. In addition, a negative treponemal test may not rule out syphilis. In a study of patients with RPR tests reactive at a dilution of greater than 1:8, antibodies to T pallidum membrane antigens were detected in a few patients with persistently negative FTA-Abs test results.(26) This finding raises the possibility that some reactions that appear to be biologically false positive may in fact represent syphilis with negative treponemal serologic tests. This study raises questions about the sensitivity of the treponemal test in diagnosing syphilis in HIV-infected patients. However, until further data are available, clinicians should consider the sensitivity of treponemal tests to be high in HIV-infected persons with untreated syphilis beyond the primary stage of infection. If asymptomatic patients have a positive nontreponemal test and negative confirmatory treponemal test results, it is very unlikely that they have active syphilis.

An increasing number of commercial and public laboratories have begun to use an enzyme immunoassay (EIA) that is specific for T pallidum, previously widely used by blood banks, as an initial serologic screening and diagnostic test for syphilis.(142-144) Initial positive tests must be reflexed to nontreponemal testing to obtain titers for the purpose of following treatment outcome. If the EIA is positive and the reflexed nontreponemal test is negative, further testing should be done using a different treponemal serologic test, given specificity concerns with the EIA test. Laboratories and localities performing these tests should also offer technical assistance and training for interpretation of results because most providers are unaccustomed to initial testing with treponemal tests. EIA-positive/nontreponemal test-negative results can occur in a number of clinical scenarios, including very early syphilis, remote syphilis that is untreated or partially treated, and with a false-positive treponemal test. To avoid missing an incipient case of syphilis when risk is present, repeating the nontreponemal test in approximately 2 to 4 weeks is recommended.

When clinical syndromes compatible with primary or secondary syphilis occur and when the findings on dark-field examinations and serologic tests are negative, clinicians should rule out the "prozone" phenomenon. This phenomenon refers to cases in which a nontreponemal serologic test is read as falsely negative because the specimen was not tested after sufficient dilution, so that the high antigen concentration did not allow detectable formation of antigen-antibody complexes.(145) When the diagnosis is in doubt, clinicians should biopsy suspicious lesions and evaluate the specimens for spirochetes using stains such as the Steiner silver stain (146) or specific DFA stains for T pallidum. Because T pallidum does not grow on artificial media, inoculation of laboratory animals (usually rabbit testicles) is the only method currently available to isolate the organism from fluids or tissues of an infected patient. This method is available only in a few research laboratories.

Clinicians should consult with an infectious diseases or laboratory medicine specialist about tests available in their areas. If biopsy specimens do not reveal spirochetes, or if techniques are not available to identify spirochetes, but clinical suspicion of syphilis remains high, clinicians may wish to presumptively treat HIV-infected patients for early syphilis and follow them closely with serial serologic testing at 1 or 2 weeks after treatment to detect any delayed antibody response.

Indications for Cerebrospinal Fluid Examination

Based on the first prospective study since the beginning of the antibiotic era (20) of patients with syphilis of less than 1 year in duration, which included follow-up for 12 months after treatment, the CDC's current STD treatment guidelines recommend CSF examination in individuals diagnosed with late latent syphilis, syphilis of unknown duration, neurologic signs or symptoms, or suspected treatment failure (Figure 1).(47) Nonetheless, this area of syphilis management remains controversial. CSF abnormalities often occur among both HIV-negative and HIV-positive patients who have primary or secondary syphilis.(20,35) Such abnormalities in HIV-infected patients are of unknown prognostic significance, and the majority respond appropriately to the currently recommended penicillin therapy.(20) Therefore, in the absence of neurologic signs or symptoms, CSF examination is currently not generally recommended before treatment in HIV-infected patients with primary, secondary, or early latent syphilis.

Based on case reports of neurosyphilis and isolation of T pallidum from the CSF of HIV-infected patients who had completed standard therapy for early syphilis, some clinicians believe that routine CSF examination should occur earlier in the course of syphilis in HIV coinfected patients, such as before treatment in all such patients (37) or 6 months after single-dose penicillin therapy for syphilis of less than 1 year in duration.

A recent multicenter study identified an increased risk of laboratory-defined, asymptomatic neurosyphilis in certain HIV-infected patients: a nearly 6-fold increase with primary, secondary, or latent syphilis and a serum titer of 1:32 or higher; and a more than 3-fold increase with syphilis and a CD4 count of 350 cells/L or lower. In patients with both of these findings, laboratory-defined neurosyphilis was more than 18 times more likely.(43) However, this study provided no longitudinal data for comparing the effectiveness of intravenous penicillin with that of intramuscular benzathine penicillin G (which is currently recommended by CDC) in patients with primary, secondary, and early latent syphilis and a titer of 1:32 or greater. A subsequent study similarly demonstrated an increased risk of laboratory-defined asymptomatic neurosyphilis in HIV-infected patients with a serum RPR titer of 1:32 or higher.(147)

Such data have prompted other experts to propose a tiered algorithm that combines CD4 count and nontreponemal test titer to identify settings in which lumbar puncture (LP) should be considered or recommended. Some experts suggest considering LP for HIV-infected patients with a nontreponemal titer greater than 1:32 independent of syphilis stage, and in those with early-stage infection and CD4 count less than 350 cells/L, regardless of titer.(148) Other experts recognize that although asymptomatic neurosyphilis is well known to be common in patients with primary, secondary, and early latent syphilis, for decades such patients--including those with serologic syphilis titers greater than 1:32 and CD4 counts less than 350 cells/L--have been treated with benzathine penicillin G, without neurologic complications in the vast majority. Thus, the benefit of treating a laboratory finding of unknown prognostic significance, at the cost of significant public health resources, is unclear.

Given these complexities, clinical judgment is required in deciding to perform LP and in diagnosing neurosyphilis. Furthermore, the clinical benefit of treating patients with laboratory-defined asymptomatic neurosyphilis has not been fully established, and the prospective study mentioned above found no benefit of such treatment at 1 year of follow-up.(20) However, the demonstration of syphilis titer and CD4 count predictors underscores the need to consider neurosyphilis in patients with high serum syphilis titers and advanced HIV infection, and to document an adequate serologic response through close follow-up of advanced HIV patients with high-titer early syphilis treated with intramuscular benzathine penicillin G in the absence of a CSF examination. Whenever the decision is made to perform an LP in a patient with syphilis, it is critical that this diagnostic procedure not interfere with timely treatment of a potentially infectious case.

Clinicians should examine the CSF for evidence of neurosyphilis in all HIV-infected patients (or patients at risk for HIV infection) who can safely undergo LP and have any unexplained behavioral abnormalities, psychological dysfunction, or ocular, auditory, or other neurologic symptoms or signs, especially those consistent with neurosyphilis. Clinicians should also examine the CSF of patients with primary, secondary, or early latent syphilis if treatment fails (eg, if the serum nontreponemal antibody titer does not decrease by 4-fold [2 dilutions] by 6-12 months for primary or secondary syphilis); if a sustained (greater than 2 weeks) 4-fold (2 dilution) or greater increase occurs; or if signs or symptoms attributable to syphilis persist, recur, or develop in the absence of reinfection.

Diagnosis of Neurosyphilis

Neurosyphilis should be considered in the differential diagnosis of neurologic disease in HIV-infected persons. The diagnosis of neurosyphilis is based on CSF findings of a reactive VDRL, presence of white blood cells, or elevated protein concentration. The VDRL is the only serologic test for syphilis approved for use on CSF specimens. If the CSF VDRL test result is negative, the diagnosis of neurosyphilis is complicated by the lack of another reliable diagnostic test and the difficulty of distinguishing between neurologic disease caused by T pallidum and that caused by HIV itself or by other CNS pathogens occurring in patients with advanced HIV disease.

Most persons with symptomatic neurosyphilis and coexisting HIV infection have a positive CSF VDRL test result and significantly elevated CSF leukocytes.(70,149) Some case reports, however, have described symptomatic neurosyphilis in HIV-infected patients whose initial CSF VDRL test results were negative, suggesting that neurosyphilis may go untreated if clinicians consider only the CSF VDRL test when making therapeutic decisions.(150) In these reported patients, the CSF VDRL test result became positive after penicillin therapy. These reports underscore the need for clinical judgment when diagnosing active neurosyphilis in HIV-infected persons and the need for better diagnostic tests.

Detection of minimally increased CSF leukocytes (5-10 cells/L) is common in HIV-infected patients in the absence of other CNS infections. In the HIV/syphilis coinfected patient, elevated CSF leukocytes of greater than 10 to 20 white blood cells/L should prompt serious consideration of a diagnosis of neurosyphilis if no other etiology is identified. The CSF FTA-Abs test may help rule out neurosyphilis if it is negative, but a positive result is nonspecific.(151)

Until improved diagnostic tests are available, clinicians may wish to initiate treatment for neurosyphilis when serologic tests for syphilis are positive, CSF leukocyte count and protein concentration are elevated but the CSF VDRL test result is negative, and all other possible etiologies have been excluded. Researchers have suggested measurement of other CSF markers, such as immunoglobulins or treponemal antigens and direct detection of treponemes,(152) but these methods have not yet been studied adequately.(153) Although detection of treponemal DNA using PCR testing is under development,(136,154,155) to date PCR has been insufficiently sensitive to aid the diagnosis of neurosyphilis.(20,137) All patients treated for neurosyphilis should be followed closely with serologic titers at 6, 12, 18, and 24 months and with repeat LPs at 6, 12, and 24 months to determine whether initial CSF abnormalities return to normal after therapy. After treatment, pleocytosis should decrease by 6 months and the CSF should be entirely normal by 2 years.


Based on available data and published consensus recommendations,(47) HIV-infected patients with primary, secondary, and early latent syphilis should be treated according to the recommendations for HIV-negative patients with primary, secondary, and early latent syphilis (Table 1). HIV-infected patients with late latent syphilis or syphilis of unknown duration should have a CSF examination before treatment and be managed accordingly.

Penicillin G regimens are recommended, whenever possible, for all stages of syphilis in HIV-infected patients (Table 1). Doxycycline and cephalosporins are not recommended by CDC's STD treatment guidelines because no studies have established an effective dose and duration for HIV-infected patients. Erythromycin and, more recently, azithromycin are not recommended. Although azithromycin once demonstrated promise for the oral treatment of incubating and early syphilis,(48,49) the recent identification of resistance (eg, as high as 56% in San Francisco in 2004) means that this drug cannot be recommended for general use in the United States.(50,51) Although amoxicillin has the theoretical benefit of penetration into the CSF and achievement of treponemicidal levels,(156,157) no difference in outcome was identified in a double-blind, randomized, controlled trial comparing benzathine penicillin G with or without amoxicillin in patients with syphilis of less than 1 year in duration.(20) In addition, no proven alternatives to penicillin are available for treating patients with neurosyphilis, congenital syphilis, or syphilis in pregnancy. Therefore, for patients with confirmed penicillin allergy, skin testing, if available, and desensitization are recommended.

Isolated case reports of treatment failure have raised questions about the efficacy of current treatment recommendations for syphilis in HIV-infected patients. Published case reports and expert opinion suggest that, in comparison with HIV-negative patients, HIV-infected patients with early syphilis may have a higher risk of neurologic complications and a higher rate of treatment failure with currently recommended regimens. The magnitude of these risks, however, although not precisely defined, is probably small. No therapies have proven more effective in preventing neurosyphilis in HIV-infected patients than the therapies recommended for HIV-uninfected patients. Careful follow-up at recommended intervals after therapy is probably the most important part of clinical management and often does not receive the consideration it deserves; greater attention is typically focused on treatment regimens and debates over when to perform LP.

To assess accurately the serologic response to treatment, providers should collect a serologic test titer on the day of treatment, especially in patients with early syphilis. Titers may rise rapidly in early infection, and failure to collect a serologic sample on the day of treatment will complicate measurement of a 4-fold decrease, especially if the previously collected titer was low.

Before administering any treatment, clinicians should warn all patients about the possibility of a Jarisch-Herxheimer reaction, which occurs most often in secondary syphilis and with penicillin therapy, and may be more common in HIV-infected patients.(20) This reaction is characterized by fever, fatigue, and transient worsening of any mucocutaneous symptoms, and usually subsides within 24 hours. These symptoms can be alleviated with acetaminophen and should not be mistaken for drug allergy. In addition, clinicians should inform HIV-infected patients that currently recommended regimens may be less effective for them than for patients without HIV infection and that close serologic follow-up is therefore essential.

Primary, Secondary, and Early Latent Syphilis in HIV-Infected Patients

Before treating, clinicians must carefully examine HIV-infected patients with primary, secondary, or early latent syphilis for clinical evidence of neurologic involvement (eg, optic and auditory symptoms, cranial nerve palsies) (Figure 2). For HIV-infected patients with incubating, primary, secondary, or early latent syphilis and with no clinical evidence of neurologic involvement, the same treatment regimen as for patients without HIV infection is recommended: 2.4 million units of benzathine penicillin G administered intramuscularly at a single session (Table 1). In addition to this regimen, some experts have recommended extra treatment (such as multiple doses of benzathine penicillin G, as recommended for latent syphilis, or other supplemental antibiotics).(37,158) However, these approaches have theoretical benefit only, with no evidence of significant clinical benefit vis-à-vis the added cost. In penicillin-sensitive patients, the clinician should confirm allergy. If compliance and close follow-up are ensured and the patient is not pregnant and refuses desensitization, the clinician may consider doxycycline (100 mg orally twice daily for 14 days). No data are available, however, on the efficacy of tetracyclines for treating syphilis in HIV-infected patients. Azithromycin is not an acceptable alternative, as discussed above. Ceftriaxone may be considered as an alternative therapy, although the optimal dose is not yet defined and close clinical and serologic follow-up is essential. If ceftriaxone is used for the treatment of early syphilis, some experts recommend 1 g daily, given intramuscularly or intravenously, for 8 to 10 days.(47,159,160) Regarding all alternatives to penicillin, if compliance and close follow-up cannot be ensured, then desensitization to penicillin and management in consultation with an infectious diseases specialist are recommended.

Late Latent Syphilis or Syphilis of Unknown Duration in HIV-Infected Patients

A careful clinical examination and CSF analysis should precede and guide the treatment of HIV-infected patients with late latent syphilis or syphilis of unknown duration (Figure 2). If the CSF examination yields no evidence of neurosyphilis, then a total of 7.2 million units of benzathine penicillin G is recommended (administered as 3 doses of 2.4 million units by intramuscular injection weekly for 3 successive weeks) (Table 1). If a CSF examination is not possible, the clinician should offer treatment for neurosyphilis; however, an LP is preferred to yield CSF markers for follow-up evaluations. In penicillin-sensitive patients, skin testing may be used to confirm allergy, after which desensitization to penicillin is recommended. If compliance and close follow-up are ensured and the patient is not pregnant and refuses desensitization, the clinician may consider doxycycline (100 mg orally twice daily for 28 days). No data are available, however, on the efficacy of tetracyclines for treating syphilis in HIV-infected patients. Azithromycin is not an acceptable alternative, as discussed above. Ceftriaxone has not been evaluated thoroughly for late latent syphilis and an optimal dose has not been defined, so close clinical and serologic follow-up is essential if this agent is used.

Tertiary Syphilis in HIV-Infected Patients

A careful clinical examination and CSF analysis should precede and guide the treatment of HIV-infected patients with tertiary syphilis. If the CSF examination yields no evidence of neurosyphilis, then a total of 7.2 million units of benzathine penicillin G is recommended (administered as 3 doses of 2.4 million units by intramuscular injection weekly for 3 successive weeks). If a CSF examination is not possible, the clinician should offer treatment for neurosyphilis; however, an LP is preferred to yield CSF markers for follow-up evaluations. In penicillin-sensitive patients, skin testing may be used to confirm allergy, after which desensitization to penicillin is recommended. Some clinicians treat all patients who have cardiovascular syphilis with a neurosyphilis regimen, regardless of the results of CSF evaluation. An infectious diseases consultation should be obtained when managing patients with tertiary syphilis.

Neurosyphilis in HIV-Infected Patients

For HIV-infected patients diagnosed with neurosyphilis (including ocular or auditory syphilis with or without positive LP results), aqueous crystalline penicillin G is the treatment of choice (administered as 18-24 million units intravenously per day; ie, 3-4 million units every 4 hours or continuous infusion for 10-14 days) (Table 1; Figure 2). The clinician should desensitize penicillin-sensitive patients; preliminary skin testing to evaluate penicillin allergy may obviate the need for desensitization. If intravenous administration is impossible, then aqueous procaine penicillin G is an alternative (administered as 2.4 million units intramuscularly daily plus probenecid 500 mg by mouth 4 times daily for 14 days). Procaine injections are painful, however, and patient compliance may be difficult to ensure.

To approximate the 21-day course of therapy for late latent disease and to address concerns about slowly dividing treponemes, most experts now recommend 3 weekly doses of benzathine penicillin G (total 7.2 million units intramuscularly) after the completion of a 14-day course of aqueous crystalline or aqueous procaine penicillin G for neurosyphilis.

No oral antibiotic alternatives are recommended for the treatment of neurosyphilis. Neither the tetracyclines nor azithromycin has been studied adequately in HIV-infected patients to support the recommendation of either for the treatment of neurosyphilis. In addition, the resistance to azithromycin recently documented precludes this option. The only alternative that has been studied is ceftriaxone 2 g intramuscularly for 14 days. Although this drug has excellent CNS penetration and a longer half-life than penicillin, the small sample size of this study precludes recommendation of this agent as standard therapy.(42) This regimen may be considered in special circumstances when neither intravenous penicillin nor procaine penicillin is feasible.

Follow-Up of HIV-Infected Patients Treated for Syphilis

Treatment failure can occur with any regimen in patients with or without HIV infection. Because the possibility of clinical relapse after syphilis therapy may be slightly higher in HIV-infected patients, the importance of closely following HIV-infected patients with syphilis cannot be overstated. Clinicians should monitor all patients carefully for persistent or recurrent symptoms, for any signs of neurologic involvement, and for increasing serologic titers.

After treatment for syphilis, patients should be reexamined clinically at 1 to 2 weeks and retested with a quantitative nontreponemal test at 3, 6, 9, 12, and 24 months after treatment for primary, secondary, and early latent syphilis and at 6, 12, 18, and 24 months after treatment for late latent syphilis or syphilis of unknown duration (Table 1; Figure 2). The objectives of such follow-up in HIV-infected patients include the following:

Documenting whether a Jarisch-Herxheimer reaction has occurred, which, when stage of infection is uncertain, may suggest earlier infection requiring epidemiologic follow-up

Documenting serologic improvement or treatment failure

Monitoring resolution, persistence, or recurrence of clinical signs and symptoms and development of any new signs or symptoms, especially those involving the CNS

Ensuring compliance with treatment and, if applicable, effective partner notification and safer-sex practices

Clinicians should monitor patients more frequently and longer if any questions exist about the adequacy of their clinical or serologic response. Clinicians must evaluate patients using the same nontreponemal test on each occasion because titers from the VDRL and RPR tests are not interchangeable.

Determining a therapeutic cure of syphilis is problematic because no simple test is available. A persistently reactive RPR or VDRL test result, typically less than 1:16 but in some cases significantly higher, may represent not an active syphilis infection but rather the serofast state, a nonpathologic, stable reactive serology. Moreover, clinical symptoms and signs of early syphilis may resolve even without treatment. Finally, serologic criteria for treatment failure have been based on curves of serologic response to treatment in patients with normal immune function.(161-163)

In a prospective study, 81% to 87% of patients with primary, secondary, or early latent syphilis had an adequate serologic response by 12 months after treatment (Table 2).(20) In the minority of patients with primary and secondary syphilis who failed to show a 4-fold decrease in nontreponemal titer after treatment, serologically defined treatment failure was more common among HIV-infected patients. This study found that although the serologic response sometimes differed between HIV-infected patients and HIV-uninfected patients after therapy for early syphilis, the difference usually was not clinically important.

The criteria for treatment failure currently include the following findings:

Persistence, recurrence, or development of clinical signs or symptoms of syphilis in the absence of reinfection

Sustained (greater than 2 weeks) 4-fold (2 dilution) increase in the nontreponemal test titer (assuming the same test type was used [eg, VDRL, RPR])

Failure of the initial nontreponemal test titer to decrease 4-fold (2 dilutions) by 6 to 12 months for primary, secondary, or early latent syphilis and by 12 to 24 months for late latent syphilis or syphilis of unknown duration (Figure 2)

Failure of nontreponemal test titers to decline 4-fold or serorevert by 12 months after therapy for primary, secondary, or early latent syphilis raises the theoretical concern that the patient has not been adequately treated for syphilis. Limited data are available to guide evaluation of the response to therapy in a patient with late latent syphilis or syphilis of unknown duration, especially if the initial nontreponemal serologic titer was low. If the nontreponemal titer fails to decline 4-fold between 12 and 24 months, especially if the initial titer was higher than 1:2, then the patient with late latent syphilis or syphilis of unknown duration may be experiencing treatment failure.

In all patients experiencing treatment failure as defined by rising titers or development of signs or symptoms of syphilis in the absence of reinfection, and in cases of possible treatment failure as defined by inadequate serologic response, especially if treated for primary, secondary, or early latent syphilis without an initial CSF examination, experts recommend CSF examination and repeat treatment with 3 weekly injections of benzathine penicillin G (2.4 million units intramuscularly every week) for a total of 7.2 million units, unless CSF examination indicates that neurosyphilis is present. Such patients should be monitored with repeat serologic testing every 6 to 12 months. Unless the patient has recurrent or new clinical signs or symptoms or a sustained, 4-fold increase in serologic titer in the absence of reinfection, most experts believe that no further therapy is needed. Those patients with persistent nontreponemal antibodies of unknown clinical significance should be monitored annually for evidence of clinical or serologic relapse. In cases of late latent syphilis or syphilis of unknown duration in which the initial titer was very low, treatment with 7.2 million units of benzathine penicillin G was documented, CSF examination was negative, and the risk of reinfection is thought to be very low, some experts would elect to monitor the titer without further evaluation and treatment and to reevaluate and treat only on the basis of recurrent signs or symptoms or a sustained, 4-fold increase in titer.

For patients with neurosyphilis, repeat serologic testing as described previously and CSF examinations at 6-month intervals are recommended until the findings have stabilized. Abnormal white blood cell count and protein level in the CSF should decrease by 6 months if no coexisting CNS infections are present, but CSF VDRL test results may remain reactive for at least 2 years. If the CSF white blood cell count is not normal or the CSF VDRL remains reactive at 2 years, and if no other cause is identified, then the patient should be reevaluated and re-treated for neurosyphilis.

Clinical Relapse Following Treatment

Mucocutaneous Syphilis

Clinicians have reported slow resolution of syphilitic skin lesions after penicillin therapy in patients with HIV infection.38 The first prospective study in the postantibiotic era examining the interval from treatment to resolution of the signs and symptoms of primary or secondary syphilis found that the resolution of chancre took somewhat longer for HIV-infected patients (median, 16.5 days) than for HIV-negative persons (median, 13 days).(20) The time to resolution of skin rash, however, did not differ by HIV status (median, 53 vs 49 days, respectively). In this study, mucocutaneous signs and symptoms of secondary syphilis recurred in 1 HIV-infected patient 6 months after treatment. It is plausible that in addition to neurologic relapse, other manifestations of syphilis could recur in a small number of HIV-infected persons after treatment.


Although neurologic relapse after penicillin therapy and optic, auditory, and neurologic signs and symptoms in early syphilis are not unique to HIV-infected patients, they are uncommon in HIV-uninfected patients.(99,100,164) Recent case reports of neurologic relapse after benzathine penicillin G and ceftriaxone therapy for syphilis in patients with HIV infection (13-16,104,165,166) have raised concerns about the adequacy of treatment regimens in HIV-infected patients with syphilis. One patient presented with eye pain, double vision, dizziness, and headache;(13) 2 weeks later, he was found in a stuporous state with a hemiparesis, homonymous hemianopsia, and expressive aphasia. CSF evaluation revealed mononuclear pleocytosis (32 cells/L), elevated protein level (92 mg/dL), and a CSF VDRL titer of 1:4. This case is consistent with meningovascular syphilis. This patient had been treated with benzathine penicillin G for secondary syphilis 5 months before this neurologic event, and the serum VDRL titer had decreased from 1:256 to 1:16. Although the serum VDRL titer around the time of the stroke was 1:256, careful contact tracing and close follow-up after the initial treatment suggested that reinfection had not occurred.

Meningovascular neurosyphilis also was diagnosed in a patient within 6 months after treatment with ceftriaxone, 2 g intravenously daily for 14 days, for latent syphilis.(165) Another patient developed syphilitic ophthalmitis with irreversible blindness 96 hours after starting high-dose intravenous penicillin treatment for syphilitic meningitis.(14) Before her treatment for syphilitic meningitis, she had received 7.2 million units of benzathine penicillin G for early syphilis. Twenty-four weeks after completing the 10-day course of 24 million units of crystalline penicillin G daily, she developed meningovascular neurosyphilis. In a prospective study of more than 500 patients with early syphilis, however, no cases of symptomatic neurosyphilis were identified during the 12-month follow-up.(20)

A follow-up study of patients treated for neurosyphilis showed that HIV-infected patients had slower resolution of serum RPR and CSF VDRL tests after treatment.(44) Most experts believe that in comparison with HIV-negative patients, HIV-infected patients with syphilis may have an increased risk of neurologic complications and higher rates of treatment failure with currently recommended regimens; the magnitude of these risks is uncertain but probably low.(47) Although advanced HIV disease is typical in these cases, specific factors associated with neurologic complications or treatment failure in these few individuals are currently unknown, and perhaps rep