Anda di halaman 1dari 3

E D I T O R I A L C O M M E N TA R Y

Neurosyphilis: Diagnosis and Response to Treatment


Daniel M. Musher
Medical Care Line, Infectious Disease Section, Michael E. DeBakey Veterans Affairs Medical Center, and Departments of Medicine and Molecular Virology
and Microbiology, Baylor College of Medicine, Houston, Texas

(See the article by Marra et al. on pages 8939)

In this issue of Clinical Infectious Diseases, 3], neurosyphilis was diagnosed clinically; CSF abnormalities or neurologic abnor-
an important article examines the sero- the diagnosis was supported by serologic malities for which other causes had been
logical response to treatment of neurosy- test results positive for syphilis and detec- excluded. Only 57% of the patients in the
philis. Marra et al. [1] show that nor- tion in CSF of antibody to cardiolipin (ini- study by Hooshmand et al. [4] had a pos-
malization of the results of the serum tially by the Wassermann reaction, then by itive CSF VDRL test result. This article [4]
antibody test for cardiolipin (rapid plasma the Hahn and other more sensitive mod- is often cited to support the notion that
reagin [RPR]) is a strong indicator of suc- ifications, and ultimately by the venereal a reactive CSF VDRL test is not a regular
cess after treatment of neurosyphilis. Most disease research laboratory [VDRL] test). feature of neurosyphilis.
of their patients had HIV infection, but If the CSF VDRL test result was negative, However, MHA-TP testing of CSF sam-
their findings are likely to apply to HIV- increased WBC count or protein concen- ples is not accepted as a diagnostic tool,
uninfected persons, as well. tration in CSF provided laboratory sup- because it is overly sensitive; passive dif-
This article [1] provides great practical port; except in forms of neurosyphilis that fusion of plasma proteins with positive se-
help to physicians who treat patients for are now rare, this was decidedly uncom- rum MHA-TP yields a positive result even
sexually transmitted diseases. Performing mon [2]. Asymptomatic neurosyphilis was when neurosyphilis is not present [5].
a lumbar puncture in a clinical setting is diagnosed on the basis of CSF VDRL test Some European authorities [6] use the
logistically difficult, and furthermore, pa- results, although in an occasional case, CSF MHA-TP assay, but they report the
tients often refuse it. Thus, it is reassuring other CSF abnormalities, in addition to a result after calculating the ratios of CSF to
to learn that normalization of the serum high serum RPR titer, might have been serum protein concentration and CSF to
RPR titer is highly predictive of a good regarded as diagnostic. Simpy stated, a di- serum MHA-TP titer to determine
response to therapy, even though this find- agnosis of neurosyphilis or the exclusion whether its detection reflects passive dif-
ing is somewhat less likely to apply to un- of this diagnosis depended largely on the fusion from plasma or local synthesis of
treated patients with AIDS. CSF VDRL test result. antibody in the CNS. Hooshmand et al.
A full understanding of the article by In 1972, Hooshmand et al. [4] reported [4] stated that 100% of their patients had
Marra et al. [1] requires further discussion a series of cases in which they diagnosed positive CSF fluorescent treponemal an-
of 2 important issuesone relating to the neurosyphilis on the basis of (1) suggestive tibody-absorption test results, as if to as-
diagnosis of neurosyphilis and the other neurologic findings, in addition to a pos- sure the reader of the correctness of their
to treatment. In the prepenicillin era [2, itive serum fluorescent treponemal anti- diagnoses. In fact, this has never been a
body-absorption test result (this highly valid basis for diagnosis of neurosyphilis,
Received 11 June 2008; accepted 19 June 2008; sensitive test, now replaced by the equiv- and it continues to astonish me that this
electronically published 20 August 2008.
Reprints or correspondence: Dr. Daniel M. Musher,
alently sensitive microhemagglutination article [4] was ever published in that form.
Infectious Disease Section, Rm. 4B-370, Michael E. DeBakey Treponema pallidum [MHA-TP] test, de- If the authors overdiagnosed neurosy-
Veterans Affairs Medical Center, 2002 Holcombe Blvd.,
tects antibody to outer cell wall proteins philis, which I believe they most certainly
Houston, TX 77030 (daniel.musher@med.va.gov).
Clinical Infectious Diseases 2008; 47:9002
of T. pallidum, and once the result is pos- did, the true percentage of patients with
This article is in the public domain, and no copyright is itive, it remains so for life) or (2) a positive negative CSF VDRL results should be
claimed.
1058-4838/2008/4707-0006
CSF fluorescent treponemal antibody-ab- much lower.
DOI: 10.1086/591535 sorption test result in addition to other There are other reasons to be suspicious

900 CID 2008:47 (1 October) EDITORIAL COMMENTARY


of the conclusions of Hooshmand et al. had reactive CSF VRDL tests, as well. derly man, now somewhat demented. The
[4]. For example, they state that 25% of However, I remain concerned about ov- pretest probability that he has neurosy-
their patients presented with a seizure. erdiagnosis of neurosyphilis in case series philis is regarded as low. Nevertheless, he
Classic treatises on syphilis [2, 3, 7, 8], in which !60% of all patients had a pos- has a reactive serum RPR test (1:1 dilu-
none of which is cited by Hooshmand et itive CSF VDRL result [1, 4]. If true cases tion), a positive serum MHA-TP test re-
al. [4], state that, except in advanced pa- of neurosyphilis were combined with cases sult, and a negative HIV ELISA result.
resis or some particularly rare forms of that did not involve neurologic disease, the Ageing alone may lead to dementia and
syphilitic meningitis, seizures are uncom- normalization of serologic measures may cause low-level reactivity of the RPR test,
mon in neurosyphilis (I do not recall hav- not be as reliable in the context of proven and the MHA-TP test result may be the
ing seen seizures in an adult that were neurosyphilis. remaining vestige of youthful ardor and/
attributable to neurosyphilis). In sum- The second topic that is worthy of dis- or indiscretion.
mary, in the vast literature that preceded cussion is treatment. Marra and colleagues May this patient have neurosyphilis? Of
the HIV era, the consensus was that CSF state that benzathine penicillin G is not course. Is he likely to? No. Does he need
VDRL test results were positive in the great recommended for persons with neurosy- a spinal tap? Even if a spinal tap is per-
majority of cases of neurosyphilis, and philis, because it yields penicillin concen- formed and the CSF sample is normal,
negative CSF VDRL test results, except in trations in the CSF that are too low to kill those who believe that the CSF VDRL test
the case of certain currently rare forms of T. pallidum [1, p. 893]. Through the result is positive in only one-half of cases
neurosyphilis, generally opposed the 1960s, official Centers for Disease Control will not have excluded the diagnosis of
diagnosis. and Prevention recommendations for neurosyphilis. What treatment should be
Shortly before the recognition of AIDS, treating neurosyphilis included 3 intra- given? Two weeks of intravenous penicillin
a number of cases of acute syphilitic men- muscular doses of benzathine penicillin requires inpatient therapy in the medical
ingitis were reported in young adults, (2.4 million U each) at weekly intervals service. On the basis of all of the previous
many of whom had recently received treat- [12]. These or similar doses [13] pre- observations, I believe that such patients
ment with benzathine penicillin. Ehrlich vented progression of asymptomatic dis- can be treated with 3 doses of benzathine
had described neurorecurrence, in ease and eradicated active disease, al- penicillin (2.4 million U) at weekly inter-
which neurosyphilis appeared in young though continued evolution of neurologic vals and that lumbar puncture probably
adults, often within a year after they re- abnormalities might occur because of does not need to be performed. I tried to
ceived inadequate therapy (cited in [7]). CNS damage [13]. maintain this position during my 2 terms
Merritt et al. [2] placed 5% of all neu- In the late 1970s and throughout the on the Center for Disease Control and
rosyphilis cases into this category. My col- 1980s, great attention was given to studies Preventions committee to formulate rec-
leagues and I [9, 10] postulated that in- that revealed that, during therapy with ommendations for sexually transmitted
adequate treatment in an immunologically benzathine penicillin, drug levels were disease therapy, but I was soundly out-
normal host was analogous to adequate generally undetectable in the CSF [14, 15]. voted by my colleagues.
treatment in a greatly immunosuppressed This finding was expected, because serum A brief anecdote raises a poignant irony.
host. In retrospect, these patients were levels do not exceed 0.1 mg/mL and CSF My colleagues and I were finishing an ar-
likely to have been infected with HIV be- levels are only a small percentage of serum ticle that we had worked on for nearly 2
fore HIV infection or AIDS was levels. Because of reports of neurosyphilis years; the article was about the clinical
recognized. appearing after treatment with benzathine manifestations of neurosyphilis in patients
HIV-infected patients with early neu- penicillin and the failure to recognize that with AIDS [9]. We wanted the great expert
rosyphilis are likely to present with in- this was a host problem, rather than an Dr. Rudolph Kampmeier to read the ar-
volvement of 1 cranial nerve, without antibiotic problem, some authorities con- ticle before we submitted it, and especially
other manifestations [9], and this form of cluded that only large doses of intravenous for that reason, we continued to polish
syphilis is more likely to be associated with penicillin could be relied on to treat neu- and refine the manuscript. When I finally
a nonreactive CSF VDRL test. Only ap- rosyphilis. The irony is that 2 weeks of called Dr. Kampmeiers office to inform
proximately three-quarters of the patients treatment with 24 million U per day of him that I was sending the article, his sec-
reported up to 1990, including many who intravenous penicillin is not needed to retary told me that he would be unable to
had only cranial nerve abnormalities, had cure neurosyphilis in the absence of HIV read it because he had just experienced a
reactive CSF VDRL tests [9], and all of the infection but may still fail to do so in pa- major stroke. Thus, we are left to interpret
patients in a small study involving indi- tients who are infected with HIV [16]. the older literature without guidance from
viduals with asymptomatic neurosyphilis What difference does all of this make? those who wrote it.
that was performed by Dowell et al. [11] Consider the hypothetical case of an el- The article by Marra et al. [1] indicates

EDITORIAL COMMENTARY CID 2008:47 (1 October) 901


that a serological response, manifested by titer predicts normalization of cerebrospinal human immunodeficiency virus (HIV) infec-
fluid and clinical abnormalities after treatment tion on the course of syphilis and on the re-
a normalization of the serum RPR reac- of neurosyphilis. Clin Infect Dis 2008; 47: sponse to treatment. Ann Intern Med 1990;
tion, is a reliable predictor of a cure after 8939 (in this issue). 113:87281.
treatment of neurosyphilis. Despite my 2. Merritt HH, Adams RD, Solomon HC. Neu- 10. Musher DM. Syphilis, neurosyphilis, penicil-
rosyphilis. Oxford, UK: Oxford University lin, and AIDS. J Infect Dis 1991; 163:12016.
reservations about whether all of the pa-
Press, 1946. 11. Dowell ME, Ross PG, Musher DM, Cate TR,
tients included in that series actually had 3. Moore JE. The modern treatment of syphilis. Baughn RE. Response of latent syphilis or neu-
neurosyphilis, a sufficient number most 2nd ed. Springfield, IL: Charles C. Thomas, rosyphilis to ceftriaxone therapy in persons
certainly did, and I feel quite confident 1941. infected with human immunodeficiency virus.
4. Hooshmand H, Escobar MR, Kopf SW. Neu- Am J Med 1992; 93:4818.
that the conclusion is valid. Much still re- rosyphilis: a study of 241 patients. JAMA 12. Syphilis: modern diagnosis and management.
mains to be learned about the diagnosis 1972; 219:7269. US Department of Health, Education and
and treatment of neurosyphilis, a fasci- 5. Jaffe HW, Kabins SA. Examination of cere- Welfare, Public Health Service, 1961.
brospinal fluid in patients with syphilis. Rev 13. Kofman O. The changing pattern of neuro-
nating and complex disease. Young inves- Infec Dis 1982; 4(Suppl):S8427. syphilis. Canad Med Assoc J 1956; 74:80712.
tigators have a fertile, albeit rocky, field if 6. Prange HW, Moskophidis M, Schipper HI, 14. Mohr JA, Griffiths W, Jackson R, Saadah H,
they choose to till it. Muller F. Relationship between neurological Bird P, Riddle J. Neurosyphilis and penicillin
features and intrathecal synthesis of IgG an- levels in cerebrospinal fluid. JAMA 1976; 236:
Acknowledgments tibodies to Treponema pallidum in untreated 22089.
and treated human neurosyphilis. J Neurol 15. Polnikorn N, Witoonpanich R, Vorachit M,
Potential conflicts of interest. D.M.M.: no 1983; 230:24152. Vejjajiva S, Vejjajiva A. Penicillin concentra-
conflicts. 7. Stokes JH. Modern clinical syphilology: di- tions in the cerebrospinal fluid after benza-
agnosis-management-case study. Philadelphia: thine penicillin and probenecid in the treat-
References Saunders, 1934. ment of syphilis. Br J Vener Dis 1982; 58:342.
8. Kampmeier RH. Essentials of syphilology. 16. Musher DM, Baughn RE. Neurosyphilis in
1. Marra CM, Maxwell CL, Tantalo LC, Sahi SK, Philadelphia: Lippincott, 1943. HIV-infected persons. N Engl J Med 1994;
Lukehart SA. Normalization of serum RPR 9. Musher DM, Hamill RJ, Baughn RE. Effect of 331:15167.

902 CID 2008:47 (1 October) EDITORIAL COMMENTARY

Anda mungkin juga menyukai