Clinical Review & Education

Review

Treatment of Syphilis
A Systematic Review
Meredith E. Clement, MD; N. Lance Okeke, MD; Charles B. Hicks, MD

Related articles pages 1922

IMPORTANCE The incidence of syphilis in the United States is increasing; it is estimated that and 1935
more than 55 000 new infections will occur in 2014. Treatment regimens are controversial, Supplemental content at
especially in specific populations, and assessing treatment response based on serology jama.com
remains a challenge.

OBJECTIVE To review evidence regarding penicillin and nonpenicillin regimens, implications

of the “serofast state,” and treatment of specific populations including those with
neurosyphilis or human immunodeficiency virus (HIV) infection and pregnant women.

EVIDENCE REVIEW We searched MEDLINE for English-language human treatment studies

dating from January 1965 until July 2014. The American Heart Association classification
system was used to rate quality of evidence.

FINDINGS We included 102 articles in our review, consisting of randomized trials,

meta-analyses, and cohort studies. Case reports and small series were excluded unless they
were the only studies providing evidence for a specific treatment strategy. We included 11
randomized trials. Evidence regarding penicillin and nonpenicillin regimens was reviewed
from studies involving 11 102 patients. Data on the treatment of early syphilis support the use
of a single intramuscular injection of 2.4 million U of benzathine penicillin G, with studies
reporting 90% to 100% treatment success rates. The value of multiple-dose treatment of Author Affiliations: Division of
early syphilis is uncertain, especially in HIV-infected individuals. Less evidence is available Infectious Diseases, Department of
Medicine, Duke University Medical
regarding therapy for late and late latent syphilis. Following treatment, nontreponemal Center, Durham, North Carolina
serologic titers should decline in a stable pattern, but a significant proportion of patients may (Clement, Okeke); Divisions of
remain seropositive (the “serofast state”). Serologic response to treatment should be evident General Internal Medicine and
Infectious Diseases, University of
by 6 months in early syphilis but is generally slower (12-24 months) for latent syphilis.
California, San Diego (Hicks).
Evidence defining treatment for HIV-infected persons and for pregnant women is limited, but
Corresponding Author: Charles B.
available data support penicillin as first-line therapy. Hicks, MD, Divisions of General
Internal Medicine and Infectious
CONCLUSIONS AND RELEVANCE The mainstay of syphilis treatment is parenteral penicillin G Diseases, University of California, San
Diego, 200 W Arbor Dr, Mail Stop
despite the relatively modest clinical trial data that support its use.
8681, San Diego, CA 92103 (cbhicks
@ucsd.edu).
JAMA. 2014;312(18):1905-1917. doi:10.1001/jama.2014.13259 Section Editor: Mary McGrae
McDermott, MD, Senior Editor.

S
yphilis is a sexually transmitted infection caused by the spi- among studies and complicating the task of drawing evidence-
rochete Treponema pallidum. First described after Euro- based conclusions.
pean explorers returned from the Americas at the end of the
15th century, syphilis has been a major cause of morbidity and mor-
tality for more than 500 years. Its clinical influence was profoundly
Public Health Consequences
diminished by the introduction of penicillin in the 1940s. After de-
clining to a historic low in the year 2000, the number of syphilis cases Syphilis is an important public health problem. Timely diagnosis and
in the United States has been increasing and now exceeds 55 000 prompt treatment are important to limiting its clinical effects. Un-
new cases each year.1 Penicillin has been the treatment of choice for treated, up to one-third of patients progress to later stages of
more than half a century, but questions regarding the appropriate disease.2 Late syphilis can cause irreversible damage to the cardio-
therapeutic regimen for various stages of syphilis still exist. Be- vascular and central nervous systems, resulting in profound mor-
cause T pallidum cannot be cultured, there is no gold standard by bidity and even death. Without adequate screening and treat-
which to assess cure. Instead, indirect means (changes in titer of ment, stillbirth, neonatal death, low birth weight, prematurity, and
syphilis serology) must be used, contributing to inconsistencies congenital syphilis may affect more than half of pregnancies among

jama.com JAMA November 12, 2014 Volume 312, Number 18 1905

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

women with syphilis.3,4 In addition, genital ulcers due to syphilis have

been linked to acquisition of human immunodeficiency virus (HIV) Results
infection.5
Treatment of Syphilis
The MEDLINE search for syphilis treatment identified 418 articles,
of which 40 were included in this review. After reviewing the refer-
Epidemiology and At-Risk Populations
ence lists of these articles for additional relevant studies, we iden-
In recent years, an increasing proportion of syphilis cases in the tified 102 articles for inclusion in our review, including RCTs, meta-
United States have been diagnosed in men who have sex with men analyses, and cohort studies. In total, the review included 11
(MSM), often associated with high-risk sexual behavior and HIV randomized trials, and the evidence regarding penicillin and non-
coinfection.6 In 2013, more penicillin regimens was reviewed from studies involving 11 102 pa-
ART antiretroviral therapy than 16 000 of the total re- tients (Table 3 and Table 4).
BPG benzathine penicillin G ported cases were primary
CSF cerebrospinal fluid and secondary syphilis, the Monitoring Syphilis Treatment Response
HIV human immunodeficiency virus most transmissible stages of Both diagnosis and assessment of treatment response rely on se-
MSM men who have sex with men
infection, and three-quar- rologic tests (Box). Treponemal tests detect antibodies to specific
ters of these occurred in antigenic components of T pallidum, while nontreponemal tests de-
RCT randomized clinical trial
MSM.6 The highest rates of tect antibodies to a nonspecific cardiolipin-cholesterol-lecithin re-
VDRL Venereal Disease Research
primar y and secondar y agin antigen produced by the host in response to syphilis infection.39
Laboratory
syphilis are currently found Nontreponemal serologic tests such as the Venereal Disease Re-
in younger men (aged 20-29 years), a change since 2006, when search Laboratory (VDRL) test or the rapid plasma reagin test are
those aged 35 to 59 years were most affected. Primary and second- used to monitor treatment response because they usually corre-
ary syphilis disproportionately affect black men, whose infection rate late with disease activity. Different nontreponemal tests do not pro-
(27.9 per 100 000) is more than 5-fold higher than the rate among vide interchangeable measurements; for example, rapid plasma re-
white men (5.4 per 100 000).1 agin titers are often higher than VDRL titers. In general, a 4-fold
decline in nontreponemal titer (such as a decrease from 1:32 to 1:8)
is needed to signify treatment response. There are limitations to re-
lying on changes in serologic titers to monitor response to treat-
Stages of Syphilis
ment. For example, a recent report demonstrated that 20% of pa-
Syphilis occurs in overlapping stages, classified according to symp- tients with syphilis experienced at least a 1-dilution (2-fold) increase
toms and time since initial infection (Table 1). Appropriate staging in nontreponemal titer within 14 days of treatment, suggesting that
is important for determining infectivity and treatment duration. A the actual baseline titer used to assess response is often
diagnosis of early syphilis (primary, secondary, and early latent syphi- underestimated.40 Coinfection with HIV may also contribute to
lis) implies that T pallidum infection occurred within the previous therapeutic uncertainty because HIV-infected patients may expe-
year.9 Late syphilis represents manifestations occurring more than rience slower serologic response rates after apparently successful
1 year and even decades after initial infection. Latent syphilis refers treatment.41-44
to T pallidum infection with reactive syphilis serologic findings but
without clinical manifestations of disease. It includes both early la- The Serofast State
tent (within 1 year of infection) and late latent (1 year or more after The serofast state refers to a situation in which nontreponemal an-
infection) syphilis. tibodies decline (often adequately) after treatment but fail to com-
pletely revert to nonreactive.45 Persons with low pretreatment ti-
ters are also sometimes said to be serofast when there is minimal
(!2-fold) or no change following treatment. The serofast state may
Methods
represent a number of different phenomena, including persistent
One of us (M.E.C.) searched MEDLINE for English-language human low-level T pallidum infection, variability of host antibody response
treatment studies dating from January 1965 through July 2014. The to infection, or tissue injury due to nonsyphilitic inflammatory
initial search was limited to clinical trials, systematic reviews, and conditions.46 A significant proportion of patients remain serofast af-
meta-analyses consisting of syphilis treatment studies. We ex- ter treatment (15%-41%), although rates depend on syphilis stage,
cluded articles involving only diagnostic testing, screening, or pre- pretreatment titer, and the time point at which response is as-
vention. Reference lists of identified articles were searched for ad- sessed, all of which vary from study to study.13,47-50 Available data
ditional relevant references (eFigure in the Supplement). Case suggest that re-treatment has a modest effect on the serofast state,
reports and series were not included unless they were the only stud- with only 27% of such patients achieving serologic cure in 1 study.46
ies providing evidence for a specific treatment strategy. The Ameri-
can Heart Association classification of recommendations was used Efficacy of Parenteral Penicillin for Early Syphilis
to grade the quality of evidence (Table 2). Grade A indicates data Treatment of syphilis is based on stage of infection and on whether
from many large randomized clinical trials (RCTs); grade B, data from there is evidence of central nervous system involvement (Figure).
fewer, smaller RCTs, careful analyses of nonrandomized studies, or Treponema pallidum remains extremely susceptible to penicillin, an
observational registries; and grade C, expert consensus.29 antimicrobial agent targeting bacterial cell wall synthesis. During

1906 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

Table 1. Stages of Syphilisa

Syphilis Stage Timing Symptoms Comment

Early syphilis (occurs within 1
year of infection)
Primary 2-4 wk after exposure Painless chancre, usually on the genitalia; Chancre usually clears spontaneously but widespread
(median, 21 d) can also occur on the perineum, anus, spirochete dissemination occurs during this stage
rectum, lips, oropharynx, or hands; may be
asymptomatic.
Secondary 2-8 wk after Wide range of systemic symptoms Rash is usually truncal and maculopapular, frequently
disappearance of including rash, fever, headache, involving the palms and soles
chancre pharyngitis, and lymphadenopathy; may be Condyloma lata (grayish white warty lesions) may be
asymptomatic seen, usually on genital mucosa
Early latent Acquisition of syphilis None Diagnosis requires (within the previous year)
within the preceding documented seroconversion or ≥4-fold increase in
year nontreponemal titer, documented seroconversion of a
treponemal test, sexual exposure to a person with
early syphilis, or only sexual contact was within the
last 12 mo (sexual debut)9
Tertiary (late; includes Occurs >1 y to decades Cardiovascular system involvement Cardiovascular disease often as a result of vasculitis of
neurosyphilis [see below], after primary infection including dilated aorta, aortic the vasa vasorum of the ascending thoracic aorta
cardiovascular, and gummatous regurgitation, or carotid ostial stenosis;
disease) gummas (granulomatous lesions) form on
skin, bones, or within organs
Latent syphilis (asymptomatic
infection with normal physical
examination findings in
association with reactive
serologic findings)
Early latent See above See above
Late latent Diagnosis of syphilis >1 None Cases not meeting above criteria for early latent
y after infection or of syphilis and cases of infection at unknown time point
unknown duration are considered late latent syphilis
Neurosyphilis (infection of
central nervous system by
Treponema pallidum, further
defined as early or late
neurosyphilis)
Early neurosyphilis Occurs <1 y after initial Usually involves the meninges and May present simultaneously with symptoms of
infection, usually within vasculature: meningitis, cranial nerve primary or secondary syphilis
weeks of exposure defects, meningovascular disease, or
stroke
Late neurosyphilis Occurs >1 y after initial Usually involves the brain and spinal cord: Much less common in current period
infection, even decades disorders such as general paresis,
after primary infection dementia, or tabes dorsalis
Serofast state Following treatment None; patients have achieved clinical cure Nontreponemal antibodies decline but do not
completely revert to nonreactive
a
Adapted from Golden et al,2 Workowski et al,7 and Cohen et al.8

more than 60 years of use, there has never been a documented case lion U of BPG are around 5%.54 In 1956, Smith et al34 reported BPG
of penicillin resistance.51 The organism’s slow dividing time (30-33 efficacy in early syphilis, finding that response rates with a single BPG
hours) requires the prolonged presence of killing (treponemicidal) injection were not different from multiple injections of procaine peni-
concentrations of antimicrobial agents. Depot preparations achieve cillin with aluminum monostearate (a long-acting formulation no lon-
this goal and have thus become the mainstay of treatment based ger widely available). At 2 years, 94.5% to 100% of all patients had
on decades of experience.52,53 Nonetheless, only limited clinical trial a seronegative status, depending on stage. Similarly, Schroeter et
evidence of efficacy exists, and interpretation of data from avail- al36 showed no difference in outcomes with single-dose BPG com-
able studies is complicated by heterogeneous definitions of syphi- pared with multiple-dose penicillin regimens (11.4% vs 10.7%-
lis stage, differences in treatment regimens, and nonstandard treat- 10.9% 2-year cumulative re-treatment rates). Some reports claim
ment outcome measurements (Table 3). 10,30-35 Despite the superiority with multiple doses rather than a single injection of
limitations in published studies, the predominance of evidence, in- BPG,31,32 but these studies have been criticized for lack of a control
cluding more recent high-quality studies, supports the use of par- group or exclusion of reinfected patients based on serological re-
enteral penicillin, and it remains the treatment of choice. The qual- sponse pattern (thus overestimating success, as some of those ex-
ity of evidence for penicillin and alternate therapies, graded according cluded patients may actually have had treatment failure).55
to the American Heart Association classification system, is shown A 1997 study by Rolfs et al10 that included 541 patients remains
in Table 2. the only large RCT for syphilis therapy from the 20th century. The
For early syphilis, studies from the 1950s onward have fo- study compared a single standard dose of 2.4 million U of intramus-
cused on benzathine penicillin G (BPG) and demonstrate favorable cular BPG with an “enhanced therapy” regimen (the addition of high-
cure rates and infrequent need for re-treatment. Historical esti- dose oral amoxicillin/probenecid to BPG) and found no difference
mates for the rate of treatment failure with a single dose of 2.4 mil- in outcome between the 2 groups. In that study, serologic failure was

jama.com JAMA November 12, 2014 Volume 312, Number 18 1907

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

Table 2. Evidence for Therapy for Each Syphilis Stage

Grade of Evidence
Grade of Evidence of of Alternative
Syphilis Stage Primary Therapy Primary Therapya Alternative Therapy Therapya Follow-up
Early syphilis BPG, 2.4 million U A10-12 Doxycycline, 100 mg twice B13-19 Clinical and serologic
(primary, secondary) intramuscularly in daily for 14 d examinations at 6 and 12 mo
1 dose
Ceftriaxone, 1-2 g/d B16,20-23
intramuscularly or
intravenously for 10-14 d
Tetracycline, 100 mg orally B13,15
4 times daily for 14 d
Azithromycin, single 2-g A11,12,24-27b
dose, only if all other
options not feasible
Early latent syphilis BPG, 2.4 million U A10-12 Nontreponemal serologic tests
intramuscularly in at 6, 12, and 24 mo
1 dose
Late latent syphilis BPG, 2.4 million U C Doxycycline, 100 mg orally C
intramuscularly in 3 twice daily for 28 d
weekly doses
Tetracycline, 500 mg orally C
4 times daily for 28 d
Tertiary syphilis BPG, 2.4 million U C
(excluding intramuscularly in 3
neurosyphilis) weekly doses
Neurosyphilis Aqueous penicillin G, 18 C Procaine penicillin, 2.4 Repeat cerebrospinal fluid
million-24 million U/d million U/d analysis every 6 months until
intravenously (every 4 h or intramuscularly, plus cell count is normal if
as continuous infusion) for probenecid, 500 mg orally pleocytosis is initially present
14 d 4 times daily, for 10-14 d
Ceftriaxone, 2 g/d
intramuscularly or
intravenously for 10-14 d
HIV coinfection Same as HIV uninfected C10 Clinical and serologic
examinations at 3, 6, 9, 12, and
24 mo
BPG, 2.4 million U C28
intramuscularly in 3
weekly doses for early
syphilis
b
Abbreviations: BPG, benzathine penicillin G; HIV, human immunodeficiency virus. Grade A due to large RCTs; however, azithromycin is used only when other
a
Grade A indicates data from many large randomized clinical trials (RCTs); options are not feasible because of macrolide resistance.
grade B, data from fewer, smaller RCTs, careful analyses of non-
randomized studies, or observational registries; and grade C, expert
consensus.29

rigidly defined (requiring a 4-fold decline in nontreponemal titer by penicillin therapy may be required for late syphilis because
3 months), contributing to the relatively high failure rate in each treponemes appear to divide more slowly during later stages of
group (18% and 17%, respectively, at 6 months).10 Only 1 patient, infection, but the validity of this concept has not been
who was HIV infected, experienced clinical treatment failure (new rigorously assessed.7 Only limited data exist regarding late latent
rash) during follow-up. However, definitive conclusions are limited syphilis therapy. In 2005, Kiddugavu et al25 published a second-
because of high loss to follow-up in the trial (52% at 1 year). ary analysis of an RCT studying 818 patients (86%) who had pre-
More recently, trials comparing penicillin with nonpenicillin regi- sumptive late latent syphilis. Benzathine penicillin G, 2.4 million U
mens confirm the efficacy of a single intramuscular injection of 2.4 intramuscularly, was given as a single dose or along with oral
million U of BPG for early syphilis. In a 2005 RCT by Riedner et al11 azithromycin. Response rates were modest (cure rates
in which 328 patients were enrolled, 95% of those receiving BPG of 56%-63%). Smith et al published a small randomized pilot
achieved serologic cure, a majority of whom had high-titer, pre- study in 2004 in which 7 of 10 HIV-infected patients treated
sumptive early latent syphilis. Hook et al12 published an RCT in 2010 with procaine penicillin had an appropriate decline in titers,
that included 517 patients and compared oral azithromycin with BPG; although 2 subsequently relapsed and 3 remained in a serofast
79% of BPG recipients achieved serologic cure by 6 months. Other state.37 Most participants in this study were not taking effective
recent retrospective studies have also demonstrated high success antiretroviral therapy (ART), and the majority were presumed to
rates with single-dose BPG for early syphilis.14,15 have late latent syphilis. Another study of HIV-infected patients
not taking suppressive ART who had late latent syphilis (some
Penicillin in Late and Late Latent Syphilis with central nervous system involvement) showed that 3 weekly
Minimal high-quality evidence exists to guide the therapy injections of BPG resulted in an appropriate serologic response in
for late syphilis. It has been postulated that longer-duration only 62%.38

1908 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

Table 3. Evidence for Penicillin Treatment of Syphilis

Outcome
Study Measure and
Source Design Participants Treatment Dosing Regimens Definitions Results Comments
Rolfs et Randomized 139 patients with 276 treated with BPG BPG: 2.4 million U IM Treatment Treatment failure: 18% 1-y follow-up (52%
al,10 1997 clinical trial primary, 253 with and 265 with enhanced as 1 dose success: with usual therapy and retained in study)
secondary, and therapy Enhanced therapy: negative or ≥4- 17% with enhanced 101/541 (19%) HIV
149 with early BPG, 2.4 million U IM fold decrease in therapy at 6 mo; 15% infected
latent syphilis in 1 dose plus 2 g RPR titer with usual therapy and
(100 with history amoxicillin and 500 14% with enhanced
of syphilis) mg probenecid orally therapy at 12 mo
3 times daily for 10 d
Talwar et Retrospective 1532 patients 1386 treated with BPG, BPG: 2.4 million U IM Patients Reactive VDRL test 30-mo surveillance
al,30 1992 study with early syphilis 17 with procaine as 1 dose, 4.8 million followed up for result in 16.5% and period
with 30 mo of penicillin, and 139 with U IM weekly as 2 serial VDRL 6.6% (primary), 27.6%
follow-up: 1008 broad-spectrum doses measurements and 8.4% (secondary),
with primary, 429 antibiotics Procaine penicillin: at 2-mo and 18.9% and 11.6%
with secondary, 0.6 million U/d for intervals for first (early latent) at 6 and
and 95 with 10 d 6 mo, 3-mo 30 mo, respectively
latent (defined as Broad-spectrum intervals for the Results for secondary
<2 y of infection) antibiotics at various next year, and and early latent syphilis
dosages 6-mo intervals were better with 4.8
for the next year million U of BPG and
procaine penicillin
Fiumara,31 Retrospective 88 patients with All treated with BPG BPG: 2.4 million U IM Patients 88/88 with primary 2-y follow-up
1986 study primary and 101 weekly as 2 doses followed up until syphilis seronegative at Excluded patients
with secondary lesions or rash 12 mo; 101/101 with with ≥4-fold increase
syphilis healed and test secondary syphilis in RPR after lesions
When combined results (RPR and seronegative at 2 years had healed or after
with previous FTA-ABS) titer had decreased
series, 588 with became negative ≥4-fold, considered
primary and 623 to have been
with secondary reinfected
syphilis No control group
Durst et Prospective 40 patients with All treated with BPG BPG: 6.0 million U Patients At 12 mo: 17/40 2-y follow-up
al,32 1973 study secondary syphilis total (2.4 million U as followed up until seronegative; at 18 mo: No control group
1 dose initially, then serologic test 33/40 seronegative; at
2.4 million U IM as 1 results became 24 mo: 40/40
dose within 5 d, then negative seronegative
1.2 million U IM as 1
dose within 3-5 d)
Jefferiss Presumed 211 patients with 107 treated with benzyl Benzyl penicillin: Patients 95% with negative Followed up for 30
and retrospective seronegative penicillin, 7 with benzyl “repeated injections” followed up with VDRL test result 22 mo mo; 238/636
Willcox,33 study primary, 179 with penicillin plus BPG, 231 of 2.5 to 5 million U clinical and after treatment (37.4%) followed up
1963 seropositive with procaine penicillin BPG: unspecified dose serologic at 22-30 mo
primary, 196 with or PAM, and 291 with as 1 or 2 injections examinations No distinction
secondary, and 50 high-dose APPG ± PAM Procaine penicillin: monthly for 6 between reinfection/
with early latent 5.1-10 million U mo, quarterly for relapse, but low rates
syphilis High-dose APPG: >10 1 year, and twice of both (cumulative
million U per year for re-treatment rate of
PAM at varied dosing second year 4.42%)
of injections
Smith et Presumed 52 patients with 52 with seronegative BPG: 2.5 million U IM Patients Seronegative at 2 y 2-y follow-up
al,34 1956 retrospective seronegative primary, 67 with as 1 dose followed up after BPG: 100% with
study primary, 67 with seropositive primary, PAM: 4.8 million U in according to seronegative primary,
seropositive and 155 with secondary a single session or 2-4 percentage 96.0% with seropositive
primary, and 155 infection treated with sessions seronegative primary, and 94.5%
with secondary BPG, 2.5 million U as (negative or less with secondary syphilis
syphilis single injection than 4 Kahn Seronegative at 2 y
166 with secondary units) after PAM: 91.0% with
infection treated with single-dose and 88.3%
PAM as 1 session with divided-dose
415 with secondary regimens
infection treated with
PAM as 2-4 sessions
Sternberg Retrospective 1400 patients Sodium penicillin in 2.4 million U “Satisfactory Seronegative primary 1-y follow-up
and study with early aqueous solution or penicillin delivered progress”: no syphilis: 86% followed No distinction of
Leifer,35 syphilis: 600 isotonic sodium over 7.5 d signs of clinical up >9 mo; satisfactory penicillin type; noted
1947 (42.8%) chloride relapse and progression in 566 “patients were
seronegative serologic test (94.3%) probably treated
primary, 564 results became/ Seropositive primary with penicillins of
(40.3%) with remained syphilis: 82% followed satisfactory potency”
seropositive negative during up >9 mo; satisfactory No control group
primary, and 236 the observation progression in 507
(56.4%) with period (89.9%)
secondary syphilis “Unsatisfactory Secondary syphilis: 84%
progress”: followed up >9 mo;
relapse or satisfactory progress at
reinfection last observation in 196
(83%)
Abbreviations: APPG, aqueous procaine penicillin G; BPG, benzathine penicillin G; FTA-ABS, fluorescent treponemal antibody absorbed; HIV, human
immunodeficiency virus; IM, intramuscularly; PAM, penicillin aluminum monostearate; RPR, rapid plasma reagin; VDRL, Venereal Disease Research Laboratory.

jama.com JAMA November 12, 2014 Volume 312, Number 18 1909

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

Table 4. Evidence for Antibiotics Other Than Penicillin for Treatment of Syphilis
Outcome Measures,
Study Treatment and Dosing Definitions, and
Source Design Participants HIV Status Regimens Comments Results
Doxycycline/
tetracycline
Li and Retrospective 641 patients with Excluded HIV- 606 (94.5%) treated with Serologic response: No significant difference
Zheng,13 cohort early syphilis (13.4% infected patients BPG, 2.4 million U IM weekly negative or ≥4-fold in serologic response
2014 primary, 52.1% as 2 doses, and 35 (5.5%) decrease in RPR or rates: 91.4% of penicillin
secondary, and with doxycycline, 100 mg serofast (±1 dilution vs 82.9% of doxycycline/
34.5% early latent) twice daily for 14 d, and from baseline if RPR was tetracycline; P = .16
tetracycline, 500 mg 4 times 1:2 or 1:1 at baseline) by
daily for 14 d 6 mo
Psomas Retrospective 116 patients with 80% HIV infected 52 treated with BPG, 2.4 Serologic response: ≤1:4 Responses: penicillin,
et al,16 early syphilis million U IM weekly as 1-3 VRDL titer 39/52 (75%);
2012 doses, 49 with ceftriaxone, Relapse: 4-fold increase ceftriaxone, 38/49
1-2 g/d for 14-21 d, and 15 in VDRL or increase to (77.6%); doxycycline,
with doxycycline, 100 mg >1:4 11/15 (73.3%)
twice or 3 times daily for No significant difference
14-21 d in time to serologic
response per treatment
(log-rank test, P = .90)
Wong et Retrospective 445 primary syphilis Excluded diagnosed 420 (94.4%) treated with Serologic response: No significant difference
al,15 cases HIV-infected patients BPG, 2.4 million U IM as 1 ≥4-fold decrease in RPR in serologic response
2008 but many had dose, and 25 (5.6%) with by 6 mo, ≥8-fold by rates: 409/420 (97.4%)
unknown HIV status doxycycline/tetracycline, 12 mo, ≥16-fold by of BPG vs 25/25 (100%)
100 mg twice daily for 14 d 24 mo, or serofast (±1 of doxycycline/
dilution from baseline if tetracycline
RPR was 1:4, 1:2, or 1:1
at baseline)
Failure: none of the
above or ≥4-fold increase
in RPR at 1-6 mo
Ghanem Retrospective 1558 patients with 13.7% HIV positive in 34/87 treated with Serologic failure: 4-fold 4/73 with serologic
et al,14 case-control early syphilis BPG vs 5.9% in doxycycline/tetracycline, increase in RPR 30-400 d failure in BPG group vs
2006 doxycycline groups 100 mg twice daily for 14 d, posttreatment or lack of 0/34 in doxycycline
met inclusion criteria and 4-fold decrease in RPR group
were compared with 73 20-400 d posttreatment
randomly selected patients
treated with BPG, 2.4 million
U IM as 1 dose
Long et Prospective 96 patients with 15/96 had HIV 58 treated with BPG, 2.4 Serologic response: No significant difference
al,17 cohort syphilis; 76 returned infection million U IM as 1 dose, and 18 negative or ≥4-fold in serologic response
2006 for 1-y follow-up with doxycycline, 100 mg decrease in RPR by 12 mo rates: 94.8% of BPG vs
Syphilis stage twice daily for 14 d if Reinfection: success then 88.9% of doxycycline;
undefined penicillin allergy subsequent 4-fold odds ratio, 2.29; 95% CI,
increase in RPR from its 0.17-21.60; P = .59
lowest level or
conversion to reactive
Failure: no 4-fold decline
in RPR or failure to
convert a 1:4 or 1:2 titer
to nonreactive
No relapse group
Schroeter Prospective 586 patients with NA 100 treated with BPG, 2.4 Cure: healed clinical Cumulative re-treatment
et al,36 early syphilis million U IM as 1 dose, 101 manifestations, prompt at 24 mo: BPG, 11.4%;
1972 with PAM, 4.8 million U, 61 and permanent decrease PAM, 10.9%; APPG,
with APPG, 4.8 million U (0.6 of VDRL 10.7%; tetracycline,
million U/d for 8 d), 107 with Relapse: reappearance of 12.7%; erythromycin,
tetracycline, 30 g (3 g/d lesions with dark field or 21.3%
×10 d), and 144 with significant increase in
erythromycin, 30 g VDRL
Failure: lack of significant
response of VDRL after
1y
Ceftriaxone
Psomas Retrospective 116 patients with 80% HIV infected 52 treated with BPG, 2.4 Serologic response: ≤1:4 Responses: penicillin,
et al,16 early syphilis million U IM as 1-3 doses, 49 VRDL titer 39/52 (75%);
2012 with ceftriaxone, 1-2 g daily Relapse: 4-fold increase ceftriaxone, 38/49
for 14-21 d, and 15 with in VDRL or increase to (77.6%); doxycycline,
doxycycline, 100 mg twice or >1:4 11/15 (73.3%)
3 times daily for 14-21 d No significant difference
in time to serologic
response per treatment
(log-rank test, P = .90)

(continued)

Efficacy of Therapies Other Than Benzathine Penicillin G randomized trials exist (Table 4).10-17,20-27,31,34,36-38 Erythromycin is
for Syphilis no longer recommended based on tolerability and resistance
The evidence for nonpenicillin treatment of syphilis consists mostly concerns,56 but doxycycline, ceftriaxone, and azithromycin are still
of small, uncontrolled, retrospective studies, although a few larger, considered potential alternatives to penicillin.

1910 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

Table 4. Evidence for Antibiotics Other Than Penicillin for Treatment of Syphilis (continued)
Outcome Measures,
Study Treatment and Dosing Definitions, and
Source Design Participants HIV Status Regimens Comments Results
Spornraft- Retrospective 24 patients with All patients HIV 12 treated with ceftriaxone, Serologic response: 11/12 with ceftriaxone
Ragaller active syphilis (21 infected 1-2 g intravenously for 10-21 negative or ≥4-fold and 12/12 with penicillin
et al,20 with primary d, and 12 with high-dose decrease in VDRL had >4-fold decline in
2011 syphilis; 3 with penicillin G (8 with BPG, 2.4 VDRL titers at median of
neurosyphilis) million U/wk as 2-3 doses, 2 18.3 mo
with clemizole penicillin G, 1
million U/d for 14-21 d, and 2
with intravenous penicillin G,
3 × 10 million U/d for 21 d)
Smith et Prospective 31 patients with All patients HIV 10 treated with procaine Serologic response: Penicillin: 7/10 had
al,37 randomized asymptomatic infected; enrolled penicillin, 2.4 million U IM, ≥4-fold decrease in RPR response (2/10 had
2004 pilot syphilis randomized before availability of plus oral probenecid and 14 Relapse: ≥4-fold increase subsequent relapse),
to procaine penicillin effective with ceftriaxone, 1 g/d for 15 in RPR, persistent titer 3/10 were serofast
or ceftriaxone antiretroviral d (only 24 followed up) ≥1:64, or clinical Ceftriaxone: 10/14 with
Most thought to have therapy progression of disease response (1/14 had
late latent syphilis Nonresponse: ≤2-fold subsequent relapse),
change in RPR 2/14 were serofast, 2/14
Serofast: persistent titer failures
after treatment
Dowell Secondary 7 patients with All patients HIV 43 treated with ceftriaxone, Serologic response: ≥4- Ceftriaxone: 28 (65%)
et al,38 neurosyphilis, 6 with infected 1-2 g/d for 10-14 d, and 13 fold decrease in RPR with had response, 5 (12%)
1992 latent syphilis, and with BPG, 2.4 million U IM as no subsequent increase were serofast, 9 (21%)
30 with presumed 3 weekly doses Relapse: ≥4-fold rise in had serologic relapse, 1
latent syphilis RPR after ≥4-fold decline (2%) progressed to
Failure: ≥4-fold RPR neurosyphilis
increase without initial BPG: 8 (62%) had
response, persistent RPR response, 1 (8%) was
≥1:64, or clinical serofast, 2 (15%) had
progression relapse, and 2 (15%) had
Serofast: persistent titer failures
after treatment with no All followed up for ≥6 mo
signs of progressive
disease
Schöfer Randomized 28 patients with NA 14 treated with ceftriaxone, 1 Serological controls were All patients had ≥4-fold
et al,21 prospective early syphilis (9 g IM 4 times daily every 2 d, repeated 1, 2, 3, 6, and decrease in VDRL titer
1989 primary, 19 and 14 with penicillin G, 1 12 mo after therapy and resolution of clinical
secondary) million U/d IM for 15 d symptoms
One adverse reaction in
penicillin G group (rash);
none in ceftriaxone group
Hook et Prospective 11 patients with NA 10 treated with ceftriaxone, Results reported at 3 mo Daily treatment: 1/10
al,22 primary and 5 with 250 mg/d for 10 d, and 6 with VDRL persistently
1988 secondary syphilis with ceftriaxone, 500 mg negative, 4/10 became
every other day as 5 doses seronegative, 5/10 had
dilutions of ≥3
Every-other-day
treatment: 1/6 became
VDRL seronegative, 5/6
had dilutions of ≥2
Of 11 patients available
for follow-up at 3-23 mo,
there were no cases with
evidence of relapse
Moorthy Randomized 18 patients with NA 5 treated with ceftriaxone, 3 Cure: clearance of Single 3-g ceftriaxone
et al,23 early syphilis g as 1 dose, 5 with lesions, nonreactive dose: 3/5 had cure, 1/5
1987 ceftriaxone, 2 g/d IM as 2 VDRL had sustained response,
doses, 3 with ceftriaxone, 2 Sustained response: 1/5 had failure
g/d IM as 5 doses, and 5 with clearance of lesions, 2 g ceftriaxone in 2 daily
BPG, 2.4 million U IM as 1 ≥4-fold decrease in VDRL doses: 3/5 had cure, 2/5
dose by 3 mo and sustained at had sustained response
12 mo 2 g ceftriaxone in 5 daily
Failure: clinical doses: 3/3 had sustained
progression or ≥4-fold response
increase in VDRL BPG: 3/5 had cure, 1/5
had sustained response,
and 1 lost to follow-up
Follow-up was at 1 y
Azithromycin
Bai et Meta-analysis 790 patients with NA 3 randomized clinical trials Serologic response: No statistically
al,26 early syphilis comparing BPG vs ≥4-fold decrease in RPR significant difference
2012 azithromycin in variable measured at 3, 6, 9, and between azithromycin
dosages 12 mo and BPG treatment in
odds of cure (odds ratio,
1.04; 95% CI, 0.69-1.56)
85.7% (341/398)
response rate for
azithromycin and 85.0%
(333/392) for BPG

(continued)

jama.com JAMA November 12, 2014 Volume 312, Number 18 1911

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

Table 4. Evidence for Antibiotics Other Than Penicillin for Treatment of Syphilis (continued)
Outcome Measures,
Study Treatment and Dosing Definitions, and
Source Design Participants HIV Status Regimens Comments Results
Hook Randomized 517 patients with Excluded HIV- 237 treated with BPG, 2.4 Serologic response: Serologic response in
et al,12 clinical trialearly syphilis: 26% infected patients million U IM as 1 dose, and negative or ≥4-fold 180/232 (77.6%) with
2010 primary, 46% 232 with azithromycin, 2.0 g decrease in RPR by 6 mo azithromycin vs 186/237
secondary, 28% early orally as 1 dose (78.5%) with BPG
latent primary, 46% Increased nonserious
early latent adverse events (mostly
secondary, and 28% gastrointestinal) in
early latent azithromycin group
Bai et Meta-analysis 476 patients with NA 4 randomized clinical trials Serologic response: 95.0% (227/239)
al,27 early syphilis comparing BPG vs ≥4-fold decrease in RPR response rate for
2008 azithromycin in variable measured at 3, 6, 9, and azithromycin and 84.0%
dosages 12 mo (199/237) for BPG
Five times more
gastrointestinal adverse
effects of azithromycin vs
BPG, but results not
significant
Kiddugavu Secondary 11 patients (1.1%) 20.8% HIV infected 18% treated with BPG alone, Serologic response: When initial titer >1:4,
et al,25 analysis of with primary, 122 2.4 million U IM as 1 dose, negative or ≥4-fold higher response rate with
2005 randomized (13%) with 17% with azithromycin alone, decrease in TRUST result azithromycin alone and
clinical trial secondary or early 1 g orally as 1 dose, and 65% by 10 mo with dual treatment vs
(not latent, and 818 with dual therapy BPG alone
randomized (86%) with All treated with single dose No difference in response
to therapy) presumed late latent IM regardless of presumed rates in analysis of lower
syphilis stage of syphilis initial titers or overall
(56%-63% response
rates)
No difference in response
rates between HIV
infected and uninfected
Riedner Randomized 25 patients with 52.1% HIV infected 163 treated with Serologic response: 97.7% response in
et al,11 clinical trial primary and 303 azithromycin, 2.0 g orally as ≥4-fold decrease in RPR azithromycin group vs
2005 with high-titer latent 1 dose, and 165 with BPG, titer at 9 mo 95.0% in BPG group
syphilis 2.4 million U IM as 1 dose
Hook et Randomized 60 patients with 4% HIV infected 14 treated with BPG, 2.4 Serologic response: Response rates: 12/14
al,24 pilot early syphilis million U IM as 1 dose; 17 clinical resolution and (86%) of BPG, 16/17
2002 with single-dose negative or 4-fold (94%) of single-dose
azithromycin and 29 with decrease in RPR azithromycin, and 24/29
double-dose azithromycin, Serofast status: clinical (83%) of double-dose
2.0 g orally as 1 dose resolution with no azithromycin
change or ≤2-fold change Failure in 1 with BPG and
in RPR 1 with double-dose
Failure: new lesions or azithromycin
≥4-fold increase in RPR

Abbreviations: APPG, aqueous procaine penicillin G; BPG, benzathine penicillin aluminum monostearate; RPR, rapid plasma reagin; TRUST, toluidine red
G; HIV, human immunodeficiency virus; IM, intramuscularly; NA, not available unheated serum test; VDRL, Venereal Disease Research Laboratory.
(study conducted prior to HIV era or did not address HIV status); PAM, penicillin

Doxycycline patients with incomplete HIV suppression. Treatment failure was

Several studies, mostly small and retrospective, indicate that doxy- most often noted in HIV-infected patients with latent syphilis.37,38
cycline/tetracycline is effective treatment for early syphilis, with se- Ceftriaxone treatment of syphilis requires multiple daily parenteral
rologic response rates of 83% to 100%.13-19 Most studies used oral doses and is both more cumbersome and more expensive than
doxycycline, 100 mg twice daily for 14 days, a regimen that has been single-dose BPG but has the advantage of treating other coexisting
shown to have 73% to 89% response rates in HIV-infected sexually transmitted infections and may be useful in patients with
patients.16,17 Doxycycline is advantageous in that it has simultane- penicillin allergy.57 Use of cephalosporins in patients with penicillin
ous activity against other sexually transmitted infections. How- allergy is discussed below.
ever, the requirement for multiple days of treatment when using
doxycycline introduces adherence concerns compared with the pre- Azithromycin
ferred single observed dosing with intramuscular BPG. As a result, Randomized clinical trials have shown that oral azithromycin (as a
doxycycline is considered an alternative to BPG in the treatment of single 1- to 2-g dose) is effective in the treatment of early syphilis,
syphilis. with much of the data derived from studies done in Africa.11,12,24 The
emergence of azithromycin resistance mutations in T pallidum with
Ceftriaxone resultant treatment failure has limited its usefulness in many re-
Several small studies of intramuscular ceftriaxone for early syphilis gions of the United States.56,58-62 In a study of 141 T pallidum samples
have shown efficacy comparable with penicillin G.16,20-23 Most of collected from 11 US clinics between 2007 and 2009, the 23s rRNA
these trials used parenteral once-daily doses of 1 to 2 g for 10 to 15 A2058G point mutation (associated with macrolide resistance/
days with serologic response rates of 65% to 100%. The lowest re- treatment failure) was found in 53% of specimens.61 Azithromycin
sponse rates were seen in HIV-infected patients, particularly among has also been associated with a 5-fold greater risk of gastrointesti-

1912 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

Figure. Suggested Algorithm for Treatment of Syphilis

Patient with presentation of syphilis

Yes No
Less than 1 year since infection?

Early syphilis Unknown

Yes No Yes No Yes No

Symptoms present? Symptoms present? Symptoms present?

A B
Chancre Systemic Neurologic Symptoms of early Symptoms Neurologic Symptoms Neurologic
symptoms symptoms disease and recent of late symptoms of late symptoms
possible exposure disease disease

Diagnostic CSF test Noa Go to Go to Go to Diagnostic CSF test Noa

results positive? A B C results positive?

Yes Diagnostic CSF test Yes Yes

results positive?

No
C
Primary Secondary Early Early latent Tertiary Late Late latent
syphilis syphilis neurosyphilis syphilis syphilis neurosyphilis syphilis

Treatment Treatment Treatment Treatment Treatment Treatment

BPG, 2.4 million U Aqueous penicillin G, BPG, 2.4 million U BPG, 2.4 million U Aqueous penicillin G, BPG, 2.4 million U
IM x 1 dose 18 million to 24 million IM x 1 dose IM for 3 weekly 18 million to 24 million IM for 3 weekly
U/d IV for 10 to 14 d doses (total dose, U/d IV for 10 to 14 d doses (total dose,
7.2 million U) 7.2 million U)

a
BPG indicates benzathine penicillin G; CSF, cerebrospinal fluid; IM, Some clinicians would treat patients with syphilis who have neurologic
intramuscular; IV, intravenous. symptoms for neurosyphilis despite negative diagnostic CSF test results.

cally occurring within 24 hours of treatment of early syphilis. It is seen

Box. Serologic Tests for Syphilis in 10% to 35% of patients and is usually self-limited.63 The reaction
is thought to result from release of lipoproteins, cytokines, and im-
Treponemal tests: detect antibodies to specific antigenic compo- mune complexes from killed organisms.63,64
nents of Treponema pallidum
Treponema pallidum particle agglutination Penicillin Allergy and Penicillin Desensitization
Fluorescent treponemal antibody absorbed Given the availability of nonpenicillin options for the treatment of
Enzyme immunoassay syphilis, the need for penicillin desensitization approaches seems
limited, and it is not well studied.65
Enzyme-linked immunosorbent assay
However, in some settings, penicillin is clearly the treatment of
Chemiluminescence assay
choice and desensitization is probably the preferred option in the
Nontreponemal tests: detect antibodies to nonspecific antigens pro-
following situations: (1) neurosyphilis in persons with a history of se-
duced by the host in response to treponemal infection
vere immediate hypersensitivity reaction to penicillin; (2) tertiary
Rapid plasma reagin
syphilis in all penicillin-allergic patients; (3) any stage of syphilis in
Venereal Disease Research Laboratory penicillin-allergic pregnant women; and (4) congenital syphilis in
Toluidine red unheated serum test penicillin-allergic infants.7
Unheated serum reagin
Special Circumstances: Neurosyphilis, HIV-Infected Persons,
and Pregnant Women
nal adverse effects compared with BPG.27 It is no longer recom- Neurosyphilis
mended as syphilis treatment unless special circumstances require Successful neurosyphilis treatment requires the presence of ad-
its use. equate, prolonged cerebrospinal fluid (CSF) concentrations of a
treponemicidal antimicrobial (Table 5). Benzathine penicillin G should
The Jarisch-Herxheimer Reaction not be used because it does not reliably achieve sufficient concen-
The Jarisch-Herxheimer reaction is a significant adverse event that trations in CSF. Intravenous administration of aqueous crystalline
can occur with any antibiotic treatment for syphilis but is most com- penicillin G achieves adequate CSF levels and is the treatment of
mon after penicillin. This reaction manifests with systemic symp- choice for neurosyphilis.69 Procaine penicillin injections also achieve
toms including fever, rash, malaise, headache, and myalgias, typi- treponemicidal levels in CSF, and there is evidence of clinical im-

jama.com JAMA November 12, 2014 Volume 312, Number 18 1913

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

Table 5. Parenteral Penicillin Regimens

Cerebrospinal Fluid
Drug Route Concentration Comments
Benzathine penicillin G66 Intramuscular Low Very low solubility leading to slow
release from injection site; blood
concentrations detectable up to 4
wk after injection
Aqueous procaine penicillin G67 Intramuscular High Slowly dissolving; blood levels
plateau at about 4 h then decrease
slowly over the next 15 to 20 h
Aqueous penicillin G Intramuscular or High Half-life of 42 min
(benzylpenicillin)68 intravenous

provement in some studies, but the regimen is often difficult to com- HIV-infected patients with syphilis, highlighting the importance of
plete because of the need for multiple intramuscular injections that appropriate treatment for both infections.85-87
can be painful and the requirement for adherence to oral proben-
ecid 4 times daily.70,71 HIV Infection and Neurosyphilis
Treatment of neurosyphilis in patients with significant penicil- Reports of increased risk of neurosyphilis emerged soon after rec-
lin allergy is challenging, and penicillin desensitization is probably ognition of the HIV epidemic.80,81,88,89 As with all patients, success-
the best option. Ceftriaxone (2 g/d intramuscularly or intrave- ful treatment of neurosyphilis in HIV-infected persons requires sus-
nously for 10-14 days) is an alternative, but evidence of its efficacy tained treponemicidal CSF penicillin concentrations. Some studies
is limited.72-74 Risk of cross-reactivity between penicillin and ceftri- suggest that as many as 60% of HIV-infected patients may have fail-
axone is negligible, but skin testing for β-lactam allergy and desen- ure of currently recommended neurosyphilis therapy with intrave-
sitization can be done if needed.7,57 nous penicillin G.81,90,91 Although penicillin is the preferred treat-
ment for neurosyphilis, ceftriaxone, 1 to 2 g/d intravenously for 10
HIV Coinfection days, is thought to be an effective alternative in HIV-infected pa-
Syphilis and HIV infection are strongly linked with one another. Most tients with neurosyphilis based on data from small observational
of the recent increase in syphilis cases in the United States has oc- studies.37,38,92 One prospective study demonstrated that HIV-
curred in MSM, and rates of HIV coinfection as high as 50% to 70% infected patients with low CD4 cell counts (<200/mL) were less likely
have been reported among MSM diagnosed as having primary and to clear CSF-VDRL titers after treatment, illustrating the need for im-
secondary syphilis.75 mune recovery facilitated by effective ART as part of optimal
There are few studies comparing syphilis treatment of HIV- management.93
coinfected patients with HIV-uninfected controls.76 Patients with HIV
infection may be at increased risk of serologic failure following treat- Pregnant Women
ment, although this is controversial.41,43 Some studies comparing se- Most cases of congenital syphilis result from transmission of T pal-
rologic response rates between HIV-infected and uninfected per- lidum to the fetus during early syphilis, while most adverse preg-
sons show no difference in treatment success rates.10,77 Other studies nancy outcomes occur in women treated for syphilis in the third tri-
attribute the perceived increased risk of treatment failure to a slower mester, demonstrating the importance of timely syphilis screening
decline in titer in HIV-infected individuals.42 Lower CD4 cell counts and treatment in pregnancy.94,95 Observational studies of syphilis
are associated with delayed treatment response and increased risk in pregnant women suggest that standard treatment based on stage
of serologic failure among HIV-infected patients.41,78 Data from the of syphilis at diagnosis is sufficient.96-99 A Cochrane review pub-
pre-ART era suggest a shorter latency period before progression to lished in 2010 found no syphilis treatment studies that met prede-
neurosyphilis and an increased risk of progression to neurosyphilis termined criteria for treatment group comparisons nor any that used
despite adequate treatment for early syphilis.79-81 Based on these randomly allocated groups of pregnant women.100,101
observations, some authorities recommend a longer duration of peni- Risks associated with the Jarisch-Herxheimer reaction in preg-
cillin for early syphilis in HIV-infected persons (7.2 million U of BPG nant women may be significant, including induction of early labor
given as 3 weekly 2.4 million–U doses).82 A recent retrospective study or fetal distress. Pregnant women should be warned about this po-
of patients with primary and secondary syphilis found a nonsignifi- tential outcome prior to treatment, but therapy should not be de-
cant increase in serologic response rates in HIV-infected patients who layed or withheld. Alternatives to penicillin are not recommended
received 3 weekly doses of penicillin compared with those treated because of potential fetal toxicity (doxycycline) or failure of treat-
with a single dose of BPG (serologic cure in 88% with a single dose ment to cross the placenta (azithromycin). There is limited evi-
vs 97% with 3 doses; P = .18); however, another study from 2013 dence suggesting that parenteral ceftriaxone is effective, but no con-
demonstrated no difference in BPG treatment outcome between trolled trials have been performed.7,101
single-dose treatment and multiple-dose (3 weekly injections) treat-
ment in HIV-infected patients with early syphilis.28,78 Effective ART
appears to reduce the likelihood of serologic failure as well as pro-
Discussion
gression to neurosyphilis.83,84
While HIV infection appears to affect syphilis outcomes, syphi- Based on this review of the available literature, the preferred regi-
lis also appears to affect the course of HIV infection. Increased HIV men for early syphilis, including primary, secondary, and early la-
replication and reduced CD4 cell counts have been reported among tent infection, is 2.4 million U of BPG in a single intramuscular injec-

1914 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

tion. Late and late latent infection should be treated with 3 weekly HIV-infected persons, and for pregnant women. The CDC’s guide-
injections of 2.4 million U of BPG, totaling 7.2 million U. Treatment lines for treating syphilis in the setting of HIV infection are largely
response should be assessed by repeat measurements of nontrepo- based on data from a 1997 RCT (done before the era of effective ART)
nemal serologic titers at 6 and 12 months following treatment of pri- that showed that HIV-infected patients respond comparably with
mary and secondary syphilis and at 6, 12, and 24 months for late and HIV-uninfected persons.10 However, only 69 HIV-infected persons
latent syphilis. Patients who are in a serofast state do not appear to in this study completed 6 months of follow-up and the only clinical
benefit substantially from re-treatment. Doxycycline and ceftriax- failure occurred in an HIV-infected patient. Given the increasing rate
one are also effective treatments for early syphilis when penicillin of syphilis in the MSM population, there is a need for additional high-
cannot be used, and reasonable evidence exists to support the use quality studies.
of either as alternatives to penicillin. Because of the potential for re-
sistance, azithromycin should not be used to treat patients in the
United States except when other agents are not available. Careful
Conclusions
follow-up is essential.
Treatment of neurosyphilis should be aqueous crystalline peni- After declining to historic lows at the turn of the 21st century, the
cillin G, 18 million to 24 million U/d via continuous infusion or di- incidence of syphilis has progressively increased in the United States,
vided into 6 daily doses for 10 to 14 days. Unfortunately, the data particularly in MSM. Penicillin remains the treatment of choice for
supporting this treatment regimen are modest and largely based on all stages of syphilis, with different regimens suggested based on
achievement of adequate treponemicidal CSF concentrations rather stage. Rigorous data from clinical trials to support the recom-
than demonstrated clinical efficacy. Individuals with HIV infection mended regimens are generally lacking, and most existing data are
should be treated similarly to uninfected patients based on the avail- limited by the lack of a gold standard to assess cure. Furthermore,
able evidence and because the preponderance of data on syphilis the absence of homogeneous diagnostic definitions and outcome
treatment in HIV-infected persons were compiled in the era prior to measurements among the available studies limits comparisons.102
the widespread availability of effective ART (which likely improves Larger, high-quality studies would be beneficial, especially in the dis-
syphilis treatment responses). Although evidence is limited in preg- proportionately affected HIV-infected population, but seem un-
nant women, the preferred treatment is penicillin, and pregnant likely to be carried out. Current recommendations are largely driven
women known to have penicillin allergy should be desensitized.7 by clinical experience and expert opinion because available data are
Our recommendations align with those of the Centers for Dis- limited and sometimes conflicting. The accumulated clinical expe-
eases Control and Prevention (CDC). Although a few large RCTs sup- rience suggests that the current guidelines are largely successful.
port the current CDC recommendations, they are primarily based With the preponderance of available clinical data supporting peni-
on results from early, uncontrolled studies and on decades of clini- cillin as the preferred treatment option, the use of benzathine peni-
cal experience. The best data exist for early syphilis because it en- cillin G offers a convenient, directly observed regimen that com-
compasses the majority of diagnosed cases. In contrast, evidence bines ease of administration with demonstrated antitreponemal
in support of treatment recommendations for late and late latent activity. Penicillin will likely remain the cornerstone of syphilis treat-
syphilis is quite limited, as is also true for neurosyphilis, for ment for years to come.

ARTICLE INFORMATION REFERENCES diseases treatment guidelines, 2010. MMWR

Author Contributions: Dr Clement had full access 1. Centers for Disease Control and Prevention. Recomm Rep. 2010;59(RR-12):1-110.
to all of the data in the study and takes Syphilis fact sheet. http://www.cdc.gov/std/syphilis 8. Cohen SE, Klausner JD, Engelman J, Philip S.
responsibility for the integrity of the data and the /STDFact-Syphilis-detailed.htm. Accessed January 19, Syphilis in the modern era. Infect Dis Clin North Am.
accuracy of the data analysis. 2014. 2013;27(4):705-722.
Study concept and design: All authors. 2. Golden MR, Marra CM, Holmes KK. Update on 9. Centers for Disease Control and Prevention. STD
Acquisition, analysis, or interpretation of data: syphilis. JAMA. 2003;290(11):1510-1514. Surveillance Case Definitions Effective January
Clement, Hicks. 2014. http://www.cdc.gov/std/stats
Drafting of the manuscript: Clement, Hicks. 3. Gomez GB, Kamb ML, Newman LM, et al.
Untreated maternal syphilis and adverse outcomes /CaseDefinitions-2014.pdf. Accessed August 17, 2014.
Critical revision of the manuscript for important
intellectual content: Okeke, Hicks. of pregnancy. Bull World Health Organ. 2013;91(3): 10. Rolfs RT, Joesoef MR, Hendershot EF, et al;
Statistical analysis: Clement, Hicks. 217-226. Syphilis and HIV Study Group. A randomized trial of
Administrative, technical, or material support: 4. Ingraham NR Jr. The value of penicillin alone in enhanced therapy for early syphilis in patients with
Clement. the prevention and treatment of congenital and without human immunodeficiency virus
Study supervision: Okeke, Hicks. syphilis. Acta Derm Venereol Suppl (Stockh). 1950;31 infection. N Engl J Med. 1997;337(5):307-314.

Conflict of Interest Disclosures: All authors have (suppl 24):60-87. 11. Riedner G, Rusizoka M, Todd J, et al. Single-dose
completed and submitted the ICMJE Form for 5. Tobian AA, Quinn TC. Herpes simplex virus type azithromycin vs penicillin G benzathine for the
Disclosure of Potential Conflicts of Interest. Dr 2 and syphilis infections with HIV. Curr Opin HIV AIDS. treatment of early syphilis. N Engl J Med. 2005;353
Hicks reports contracted research agreements with 2009;4(4):294-299. (12):1236-1244.
Argos, Bristol-Myers Squibb, Gilead, Janssen, 6. Patton ME, Su JR, Nelson R, Weinstock H; 12. Hook EW III, Behets F, Van Damme K, et al. A
Merck, and ViiV and consulting fees from Bristol- Centers for Disease Control and Prevention. phase III equivalence trial of azithromycin vs
Myers Squibb, Gilead, Janssen, Merck, and ViiV. No Primary and secondary syphilis—United States, benzathine penicillin for treatment of early syphilis.
other disclosures were reported. 2005-2013. MMWR Morb Mortal Wkly Rep. 2014;63 J Infect Dis. 2010;201(11):1729-1735.
Submissions:We encourage authors to submit (18):402-406. 13. Li J, Zheng H-Y. Early syphilis: serological
papers for consideration as a Review. Please 7. Workowski KA, Berman S; Centers for Disease treatment response to doxycycline/tetracycline vs
contact Mary McGrae McDermott, MD, at Control and Prevention. Sexually transmitted benzathine penicillin. J Infect Dev Ctries. 2014;8(2):
mdm608@northwestern.edu. 228-232.

jama.com JAMA November 12, 2014 Volume 312, Number 18 1915

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Clinical Review & Education Review Treatment of Syphilis

14. Ghanem KG, Erbelding EJ, Cheng WW, Rompalo 32. Durst RD Jr, Sibulkin D, Trunnell TN, Allyn B. secondary, latent, and tertiary syphilis. PLoS One.
AM. Doxycycline compared with benzathine Dose-related seroreversal in syphilis. Arch Dermatol. 2013;8(7):e70102.
penicillin for the treatment of early syphilis. Clin 1973;108(5):663-664. 49. Dionne-Odom J, Karita E, Kilembe W, et al.
Infect Dis. 2006;42(6):e45-e49. 33. Jefferiss FJG, Willcox RR. Treatment of early Syphilis treatment response among HIV-discordant
15. Wong T, Singh AE, De P. Primary syphilis: syphilis with penicillin alone. Br J Vener Dis. 1963;39 couples in Zambia and Rwanda. Clin Infect Dis.
serological treatment response to (3):143-148. 2013;56(12):1829-1837.
doxycycline/tetracycline vs benzathine penicillin. 34. Smith CA, Kamp M, Olansky S, Price EV. 50. Seña AC, Wolff M, Martin DH, et al. Predictors
Am J Med. 2008;121(10):903-908. Benzathine penicillin G in the treatment of syphilis. of serological cure and serofast state after
16. Psomas KC, Brun M, Causse A, Atoui N, Reynes Bull World Health Organ. 1956;15(6):1087-1096. treatment in HIV-negative persons with early
J, Le Moing V. Efficacy of ceftriaxone and 35. Sternberg TH, Leifer W. Treatment of early syphilis. Clin Infect Dis. 2011;53(11):1092-1099.
doxycycline in the treatment of early syphilis. Med syphilis with penicillin. JAMA. 1947;133(1):1-5. doi:10 51. Lewis DA, Lukehart SA. Antimicrobial resistance
Mal Infect. 2012;42(1):15-19. .1001/jama.1947.02880010003001. in Neisseria gonorrhoeae and Treponema pallidum.
17. Long CM, Klausner JD, Leon S, et al; NIMH 36. Schroeter AL, Lucas JB, Price EV, Falcone VH. Sex Transm Infect. 2011;87(suppl 2):ii39-ii43.
Collaborative HIV/STD Prevention Trial Group. Treatment for early syphilis and reactivity of 52. Nell EE. Comparative sensitivity of treponemes
Syphilis treatment and HIV infection in a serologic tests. JAMA. 1972;221(5):471-476. of syphilis, yaws, and bejel to penicillin in vitro, with
population-based study of persons at high risk for observations on factors affecting its treponemicidal
sexually transmitted disease/HIV infection in Lima, 37. Smith NH, Musher DM, Huang DB, et al.
Response of HIV-infected patients with action. Am J Syph Gonorrhea Vener Dis. 1954;38(2):
Peru. Sex Transm Dis. 2006;33(3):151-155. 92-106.
asymptomatic syphilis to intensive intramuscular
18. Onoda Y. Therapeutic effect of oral doxycycline therapy with ceftriaxone or procaine penicillin. Int J 53. Norris SJ, Edmondson DG. In vitro culture
on syphilis [in Japanese]. Jpn J Antibiot. 1980;33 STD AIDS. 2004;15(5):328-332. system to determine MICs and MBCs of
(1):18-28. antimicrobial agents against Treponema pallidum
38. Dowell ME, Ross PG, Musher DM, et al.
19. Harshan V, Jayakumar W. Doxycycline in early Response of latent syphilis or neurosyphilis to subsp pallidum (Nichols strain). Antimicrob Agents
syphilis: a long term follow up. Indian J Dermatol. ceftriaxone therapy in persons infected with human Chemother. 1988;32(1):68-74.
1982;27(4):119-124. immunodeficiency virus. Am J Med. 1992;93(5): 54. Idsoe O, Guthe T, Willcox RR. Penicillin in the
20. Spornraft-Ragaller P, Abraham S, Lueck C, 481-488. treatment of syphilis. Bull World Health Organ.
Meurer M. Response of HIV-infected patients with 39. Association of Public Health Laboratories. 1972;47(suppl):1-68.
syphilis to therapy with penicillin or intravenous Laboratory Diagnostic Testing for Treponema 55. Rolfs RT. Treatment of syphilis, 1993. Clin Infect
ceftriaxone. Eur J Med Res. 2011;16(2):47-51. pallidum: Expert Consultation Meeting Summary Dis. 1995;20(suppl 1):S23-S38.
21. Schöfer H, Vogt HJ, Milbradt R. Ceftriaxone for Report. January 2009. http://www.aphl.org 56. Lukehart SA, Godornes C, Molini BJ, et al.
the treatment of primary and secondary syphilis. /aphlprograms/infectious/std/Documents/ID Macrolide resistance in Treponema pallidum in the
Chemotherapy. 1989;35(2):140-145. _2009Jan_Laboratory-Guidelines-Treponema United States and Ireland. N Engl J Med. 2004;
22. Hook EW III, Roddy RE, Handsfield HH. -pallidum-Meeting-Report.pdf. Accessed October 351(2):154-158.
Ceftriaxone therapy for incubating and early 19, 2013.
57. Pichichero ME, Casey JR. Safe use of selected
syphilis. J Infect Dis. 1988;158(4):881-884. 40. Holman KM, Wolff M, Seña AC, et al. Rapid cephalosporins in penicillin-allergic patients.
23. Moorthy TT, Lee CT, Lim KB, Tan T. Ceftriaxone plasma reagin titer variation in the 2 weeks after Otolaryngol Head Neck Surg. 2007;136(3):340-347.
for treatment of primary syphilis in men. Sex syphilis therapy. Sex Transm Dis. 2012;39(8):645-647.
58. Klausner JD, Kohn RP, Kent CK. Azithromycin
Transm Dis. 1987;14(2):116-118. 41. Ghanem KG, Erbelding EJ, Wiener ZS, Rompalo vs penicillin for early syphilis. N Engl J Med. 2006;
24. Hook EW III, Martin DH, Stephens J, et al. A AM. Serological response to syphilis treatment in 354(2):203-205.
randomized, comparative pilot study of HIV-positive and HIV-negative patients attending
sexually transmitted diseases clinics. Sex Transm 59. Marra CM, Colina AP, Godornes C, et al.
azithromycin vs benzathine penicillin G for Antibiotic selection may contribute to increases in
treatment of early syphilis. Sex Transm Dis. 2002; Infect. 2007;83(2):97-101.
macrolide-resistant Treponema pallidum. J Infect Dis.
29(8):486-490. 42. Manavi K, McMillan A. The outcome of 2006;194(12):1771-1773.
25. Kiddugavu MG, Kiwanuka N, Wawer MJ, et al; treatment of early latent syphilis and syphilis with
undetermined duration in HIV-infected and 60. Van Damme K, Behets F, Ravelomanana N,
Rakai Study Group. Effectiveness of syphilis et al. Evaluation of azithromycin resistance in
treatment using azithromycin and/or benzathine HIV-uninfected patients. Int J STD AIDS. 2007;18
(12):814-818. Treponema pallidum specimens from Madagascar.
penicillin in Rakai, Uganda. Sex Transm Dis. 2005;32 Sex Transm Dis. 2009;36(12):775-776.
(1):1-6. 43. González-López JJ, Guerrero MLF, Luján R,
et al. Factors determining serologic response to 61. A2058G Prevalence Workgroup. Prevalence of
26. Bai ZG, Wang B, Yang K, et al. Azithromycin vs the 23S rRNA A2058G point mutation and
penicillin G benzathine for early syphilis. Cochrane treatment in patients with syphilis. Clin Infect Dis.
2009;49(10):1505-1511. molecular subtypes in Treponema pallidum in the
Database Syst Rev. 2012;6:CD007270. United States, 2007 to 2009. Sex Transm Dis. 2012;
27. Bai Z-G, Yang K-H, Liu Y-L, et al. Azithromycin 44. Knaute DF, Graf N, Lautenschlager S, et al. 39(10):794-798.
vs benzathine penicillin G for early syphilis. Int J STD Serological response to treatment of syphilis
according to disease stage and HIV status. Clin 62. Centers for Disease Control and Prevention.
AIDS. 2008;19(4):217-221. Azithromycin treatment failures in syphilis
Infect Dis. 2012;55(12):1615-1622.
28. Tittes J, Aichelburg MC, Antoniewicz L, Geusau infections—San Francisco, California, 2002-2003.
A. Enhanced therapy for primary and secondary 45. Ghanem KG, Workowski KA. Management of MMWR Morb Mortal Wkly Rep. 2004;53(9):197-198.
syphilis. Int J STD AIDS. 2013;24(9):703-711. adult syphilis. Clin Infect Dis. 2011;53(suppl 3):S110-
S128. 63. Yang C-J, Lee N-Y, Lin Y-H, et al.
29. Gibbons RJ, Smith S, Antman E; American Jarisch-Herxheimer reaction after penicillin therapy
College of Cardiology; American Heart Association. 46. Seña AC, Wolff M, Behets F, et al. Response to among patients with syphilis in the era of the HIV
American College of Cardiology/American Heart therapy following retreatment of serofast early infection epidemic. Clin Infect Dis. 2010;51(8):976-
Association clinical practice guidelines, I. Circulation. syphilis patients with benzathine penicillin. Clin 979.
2003;107(23):2979-2986. Infect Dis. 2013;56(3):420-422.
64. Pound MW, May DB. Proposed mechanisms
30. Talwar S, Tutakne MA, Tiwari VD. VDRL titres in 47. Agmon-Levin N, Elbirt D, Asher I, et al. Syphilis and preventative options of Jarisch-Herxheimer
early syphilis before and after treatment. Genitourin and HIV co-infection in an Israeli HIV clinic. Int J STD reactions. J Clin Pharm Ther. 2005;30(3):291-295.
Med. 1992;68(2):120-122. AIDS. 2010;21(4):249-252.
65. Cernadas JR, Brockow K, Romano A, et al;
31. Fiumara NJ. Treatment of primary and 48. Tong M-L, Lin L-R, Liu G-L, et al. Factors European Network of Drug Allergy and the EAACI
secondary syphilis. J Am Acad Dermatol. 1986;14 associated with serological cure and the serofast Interest Group on Drug Hypersensitivity. General
(3):487-491. state of HIV-negative patients with primary, considerations on rapid desensitization for drug

1916 JAMA November 12, 2014 Volume 312, Number 18 jama.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015

 Treatment of Syphilis Review Clinical Review & Education

hypersensitivity—a consensus statement. Allergy. 78. Jinno S, Anker B, Kaur P, et al. Predictors of 91. Malone JL, Wallace MR, Hendrick BB, et al.
2010;65(11):1357-1366. serological failure after treatment in HIV-infected Syphilis and neurosyphilis in a human
66. National Library of Medicine. Bicillin L-A patients with early syphilis in the emerging era of immunodeficiency virus type 1 seropositive
(penicillin G benzathine) injection. http://dailymed universal antiretroviral therapy use. BMC Infect Dis. population. Am J Med. 1995;99(1):55-63.
.nlm.nih.gov/dailymed/drugInfo.cfm?setid= 2013;13:605. 92. Marra CM, Boutin P, McArthur JC, et al. A pilot
012d46f1-d0a0-4676-a879-cd320297ab16. 79. Pialoux G, Vimont S, Moulignier A, et al. Effect study evaluating ceftriaxone and penicillin G as
Accessed May 24, 2014. of HIV infection on the course of syphilis. AIDS Rev. treatment agents for neurosyphilis in human
67. National Library of Medicine. Penicillin G 2008;10(2):85-92. immunodeficiency virus–infected individuals. Clin
procaine injection. http://dailymed.nlm.nih.gov 80. Musher DM, Hamill RJ, Baughn RE. Effect of Infect Dis. 2000;30(3):540-544.
/dailymed/drugInfo.cfm?setid=07fe2b47-34ac human immunodeficiency virus (HIV) infection on 93. Marra CM, Maxwell CL, Tantalo L, et al.
-4a20-be67-3c788c77b207#nlm34090-1. Accessed the course of syphilis and on the response to Normalization of cerebrospinal fluid abnormalities
May 24, 2014. treatment. Ann Intern Med. 1990;113(11):872-881. after neurosyphilis therapy. Clin Infect Dis. 2004;38
68. National Library of Medicine. Penicillin G 81. Gordon SM, Eaton ME, George R, et al. The (7):1001-1006.
sodium injection. http://dailymed.nlm.nih.gov response of symptomatic neurosyphilis to 94. Hawkes SJ, Gomez GB, Broutet N. Early
/dailymed/drugInfo.cfm?setid=23b6d4a3-b273 high-dose intravenous penicillin G in patients with antenatal care: does it make a difference to
-4e10-9da9-1376933fdbdf. Accessed May 24, 2014. human immunodeficiency virus infection. N Engl J outcomes of pregnancy associated with syphilis?
69. Mohr JA, Griffiths W, Jackson R, et al. Med. 1994;331(22):1469-1473. PLoS One. 2013;8(2):e56713.
Neurosyphilis and penicillin levels in cerebrospinal 82. Musher DM. Syphilis, neurosyphilis, penicillin, 95. Qin J, Yang T, Xiao S, et al. Reported estimates
fluid. JAMA. 1976;236(19):2208-2209. and AIDS. J Infect Dis. 1991;163(6):1201-1206. of adverse pregnancy outcomes among women
70. Dunlop EM, Al-Egaily SS, Houang ET. 83. Ghanem KG, Moore RD, Rompalo AM, et al. with and without syphilis. PLoS One. 2014;9(7):
Production of treponemicidal concentration of Antiretroviral therapy is associated with reduced e102203.
penicillin in cerebrospinal fluid. BMJ (Clin Res Ed). serologic failure rates for syphilis among 96. Blencowe H, Cousens S, Kamb M, et al. Lives
1981;283(6292):646. HIV-infected patients. Clin Infect Dis. 2008;47(2): Saved Tool supplement detection and treatment of
71. Hellerstrom S, Skog E. Outcome of penicillin 258-265. syphilis in pregnancy to reduce syphilis related
therapy of syphilis. Acta Derm Venereol. 1962;42: 84. Ghanem KG, Moore RD, Rompalo AM, et al. stillbirths and neonatal mortality. BMC Public Health.
179-194. Neurosyphilis in a clinical cohort of HIV-1-infected 2011;11(suppl 3):S9.

72. Gentile JH, Viviani C, Sparo MD, Arduino RC. patients. AIDS. 2008;22(10):1145-1151. 97. Alexander JM, Sheffield JS, Sanchez PJ, et al.
Syphilitic meningomyelitis treated with ceftriaxone. 85. Buchacz K, Patel P, Taylor M, et al. Syphilis Efficacy of treatment for syphilis in pregnancy.
Clin Infect Dis. 1998;26(2):528. increases HIV viral load and decreases CD4 cell Obstet Gynecol. 1999;93(1):5-8.

73. Hook EW III, Baker-Zander SA, Moskovitz BL, counts in HIV-infected patients with new syphilis 98. Watson-Jones D, Gumodoka B, Weiss H, et al.
et al. Ceftriaxone therapy for asymptomatic infections. AIDS. 2004;18(15):2075-2079. Syphilis in pregnancy in Tanzania, II. J Infect Dis.
neurosyphilis. Sex Transm Dis. 1986;13(3)(suppl): 86. Palacios R, Jiménez-Oñate F, Aguilar M, et al. 2002;186(7):948-957.
185-188. Impact of syphilis infection on HIV viral load and 99. Berman SM. Maternal syphilis. Bull World
74. Shann S, Wilson J. Treatment of neurosyphilis CD4 cell counts in HIV-infected patients. J Acquir Health Organ. 2004;82(6):433-438.
with ceftriaxone. Sex Transm Infect. 2003;79(5): Immune Defic Syndr. 2007;44(3):356-359. 100. Walker GJ. Antibiotics for syphilis diagnosed
415-416. 87. Kofoed K, Gerstoft J, Mathiesen LR, Benfield T. during pregnancy. Cochrane Database Syst Rev.
75. Su J, Weinstock H. Epidemiology of co-infection Syphilis and human immunodeficiency virus (HIV)-1 2001;(3):CD001143.
with HIV and syphilis in 34 states, United coinfection. Sex Transm Dis. 2006;33(3):143-148. 101. Zhou P, Gu Z, Xu J, et al. A study evaluating
States—2009. In: Proceedings of the 2011 National 88. Flood JM, Weinstock HS, Guroy ME, et al. ceftriaxone as a treatment agent for primary and
HIV Prevention Conference; August 14-17, 2011; Neurosyphilis during the AIDS epidemic, San secondary syphilis in pregnancy. Sex Transm Dis.
Atlanta, GA. Francisco, 1985-1992. J Infect Dis. 1998;177(4):931- 2005;32(8):495-498.
76. Blank LJ, Rompalo AM, Erbelding EJ, et al. 940. 102. Parkes R, Renton A, Meheus A,
Treatment of syphilis in HIV-infected subjects. Sex 89. Schöfer H, Imhof M, Thoma-Greber E, et al; Laukamm-Josten U. Review of current evidence
Transm Infect. 2011;87(1):9-16. German AIDS Study Group. Active syphilis in HIV and comparison of guidelines for effective syphilis
77. Goeman J, Kivuvu M, Nzila N, et al. Similar infection. Genitourin Med. 1996;72(3):176-181. treatment in Europe. Int J STD AIDS. 2004;15(2):73-
serological response to conventional therapy for 90. Marra CM, Longstreth WT Jr, Maxwell CL, 88.
syphilis among HIV-positive and HIV-negative Lukehart SA. Resolution of serum and cerebrospinal
women. Genitourin Med. 1995;71(5):275-279. fluid abnormalities after treatment of neurosyphilis.
Sex Transm Dis. 1996;23(3):184-189.

jama.com JAMA November 12, 2014 Volume 312, Number 18 1917

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Sanofi-Aventis Recherche& Development User on 01/23/2015