close

Вход

Забыли?

вход по аккаунту

?

Clinical significance of cerebrospinal fluid tests for neurosyphilis.

код для вставкиСкачать
Clinical Signrficance of
Cerebrospinal Fluid Tests for Neurosyphhs
Larry E. Davis, MD, and Joseph W. Schmitt, MD
From 1978 to 1987, 1,665 cerebrospinal fluid (CSF) fluorescent treponemal antibody absorption (CSF-FTA-ABS)tests
were performed as the screening procedure for neurosyphilis. The CSF samples from 48 patients were reactive, and the
medical history and results of the CSF-Venereal Disease Research Laboratory test (CSF-VDRL) for syphilis for 38 of
these patients were reviewed. Likely active neurosyphilis was diagnosed if the patient had a reactive CSF-FTA-ABS
test, recent onset of neurological signs consistent with neurosyphilis, abnormal CSF, and no other recognized cause for
the neurological illness. Fifteen patients were so classified. Four had a reactive CSF-VDRL test. The specificity of the
CSF-VDRL in diagnosing likely active neurosyphilis was loo%, but the sensitivity was only 27%. The insensitivity of
the CSF-VDRL test limits its usefulness as a screening test for neurosyphilis. The CSF-FTA-ABS test appears more
sensitive for screening but is less specific than the CSF-VDRL test in distinguishing currently active neurosyphilis from
past syphilis. These findings imply that clinical judgment is still essential in establishing the diagnosis of active
neurosyphilis.
Davis LE, Schmitt JW. Clinical significance of cerebrospinal fluid tests
for neurosyphilis. Ann Neurol 1989;25:50-55
The Centers for Disease Control (CDC) recommends
use of the cerebrospinal fluid (CSF)-Venereal Disease
Research Laboratory (CSF-VDRL) test in establishing
the diagnosis of neurosyphilis when serological tests
are reactive and the patient’s history is unknown 111.
Whereas a reactive CSF-VDRL test has been widely
accepted as serological evidence of neurosyphilis,
there has been some concern that the test is not sensitive enough to detect all cases of neurosyphilis 11, 21.
It is currently unknown how often active neurosyphilis
is associated with a nonreactive CSF-VDRL test. The
estimates have varied based on what criteria have been
used to diagnose neurosyphilis. If clinical signs and
symptoms compatible with neurosyphilis are used as
evidence of neurosyphilis, the number of falsenegative CSF-VDRL tests may be over 40% 13-51.
Unfortunately, the signs of neurosyphilis are not
pathognomonic and often overlap with signs of other
diseases 161, limiting the ability to use solely the clinical picture to make a diagnosis. The fluorescent treponemal antibody absorption (FTA-ABS) test has been
shown to be both sensitive and specific for syphilis 171.
Its use to diagnose neurosyphilis in the CSF has been
controversial 18, 91. Much of this controversy stems
from concern that the test may be too sensitive [lo],
and use of a reactive CSF-FTA-ABS test to diagnose
neurosyphilis has been discouraged by the CDC {8,97
because it may overdiagnose the disease.
We used the sensitivity of the CSF-FTA-ABS test to
identify individuals who were candidates for having
active neurosyphilis. Since 1978, the CSF-FTA-ABS
test has been the CSF screening test for neurosyphilis
at the University of New Mexico Hospital. Samples
reactive in the CSF-FTA-ABS test undergo a CSFVDRL test. This unusual sequence of tests enabled us
to review the medical records of individuals who had
reactive CSF-FTA-ABS tests as an approach to determining the sensitivity and specificity of the CSF-VDRL
test.
From the Neurology Service, Veterans Administration Medical
Center, Albuquerque, and the Departments of Neurology and Microbiolom, University of New Mexico School of Medicine, A h querque, NM.
Received Jan 14, 1988, and in revised form Jun 27. Accepted for
publication Jun 27, 1988.
Methods
The syphilis serological records from the University of New
Mexico Medical Center from January 1978 to September
1987 were reviewed. Cases of possible congenital syphilis
were not reviewed. The CSF syphilis serological test was
ordered on patients undergoing a lumbar puncture in whom
neurosyphilis was considered a possible diagnosis. All CSF
specimens underwent a CSF-FTA-ABS test as the screening
procedure. For most reactive CSF samples, a CSF-VDRL test
was then performed on the same sample. A retrospective
chart review of patients with reactive CSF-FTA-ABS tests
was then undertaken.
For the purpose of this study, the diagnosis of neurosyphilis was divided into two categories:likely active neurosyphilis
and unlikely active neurosyphilis.The four criteria for likely
active neurosyphilis included (1) the recent onset (< 6
months) or progression of one or more of the following
~~
50
~~~
Address correspon~enceto D~~
~~~~~l~~
~
service,
i
~ VA Med,
ical Center, 2100 Ridgecrest Drive SE, Albuquerque, NM 87108.
clinical signs: decreased cognition, unexplained personality
change, seizures, hemiparesis, cranial nerve palsies, optic
nerve dysfunction, and chorioretinitis; (2) CSF containing
more than 40 mg/dl glucose, sterile routine bacterial and
fungal cultures plus more than 5 white blood cells per cubic
millimeter or more than 45 mg/dl protein or both; ( 3 ) a
reactive CSF-FTA-ABS test and a reactive serum ETA-ABS
test; and (4) no other recognized cause for the neurological
illness. The type of neurosyphilis these patients had was determined by standard criteria I1 1, 12’1. Patients’ afflictions
were categorized as unlikely active neurosyphilis if the patient had a reactive CSF-FTA-ABS test but did not fulfill the
above criteria.
The CSF-FTA-ABS test was performed on fresh CSF using the CDC guidelines [ 131and commercial reagents (Zeus
Scientific Corporation, Raritan, NJ). The CSF was diluted
1:5 in the sorbent. All borderline and 1+ reactive tests were
repeated. Tests evaluated as 1 + twice were reported as reactive. The CSF-VDRI,test was performed on fresh CSF using
CDC guidelines I131 and commercial reagents (Fisher
ScientificLaboratory, Orangeburg, NY) at the New Mexico
State Reference Serology Laboratory.
Results
CSF-FTA-ABS tests were performed on 1,665 samples. Forty-eight (3%) were reactive. Medical records
on 45 patients were available for review. Seven patients were not included in the study because no CSFVDRL test was performed. Tables 1 and 2 give the
clinical, CSF, and syphilis serological findings on the
remaining 38 patients. Four patients had reactive CSFVDRL tests. Fifteen patients were thought likely to
have active neurosyphilis, and 23 patients unlikely to
have active neurosyphilis. Serum FTA-ABS test results were available for 18 of the 23 cases classified as
unlikely active neurosyphilis and 15 of the 15 cases
classified as likely active neurosyphilis. All of these
tests were reactive.
Eleven patients had neurological signs of general
paresis, 1 of meningovascular syphilis, 1 of acute syphilitic meningitis, and 2 of both general paresis and
meningovascular syphilis together. The mean age of
patients with likely active neurosyphilis was 60 years,
compared with 56 years for those unlikely to have
active neurosyphilis. Forty percent of patients with
likely active neurosyphilis had previously diagnosed
syphilis, compared with 43% of those unlikely to have
active neurosyphilis. All patients with likely active
neurosyphilis either were not previously treated for
syphilis or had been treated with antibiotics at least 1
year before the lumbar puncture. In most cases, patients could not recall what antibiotic they had received. Follow-up neurological evaluations were seldom available, since most patients failed to show up
for their follow-up clinic appointments. When information was available, the clinical signs had improved or
shown no further progression upon follow-up. Patients
classified as unlikely to have active neurosyphilis were
thought most likely to have had either asymptomatic
neurosyphilis, treated neurosyphilis, or syphilis other
than neurosyphilis that was active or treated in which
their CSF-FTA-ABS test was thought to be falsepositive due to serum contamination.
Three percent of all CSF-FTA-ABS screening tests
were reactive. However, only 39% of the patients with
reactive CSF-FTA-ABS tests were thought to have
likely active neurosyphilis. The specificity of the CSFVDRL test for likely active neurosyphilis was loo%,
but the sensitivity for active disease was only 27%.
The predictive value of a nonreactive CSF-VDRL in
the setting of a reactive CSF-FTA-ABS test was 68%.
Discussion
To diagnose syphilis, one wants the most sensitive and
most specific test available to detect any Treponema
pallidam antibody in the blood. To diagnose neurosyphilis, one wants a CSF test that can distinguish between T . pallidtm antibody that passively enters the
CSF through the choroid plexus and T . pallidam antibody that is locally produced in the brain and meninges by B cells that have migrated into the central nervous system in response to a local spirochetal infection.
Therefore, an extremely sensitive serological test
might not be helpful in diagnosing neurosyphilis.
In previous years patients carrying the diagnosis of
active neurosyphilis received the same amount of penicillin as any patient with syphilis of over 1 year’s duration [l]. It is now recognized that the dose of penicillin
was occasionally inadequate and failed to cure the central nervous system disease [141. In 1985 the CDC
suggested that patients with neurosyphilis be treated
longer and with larger total doses of penicillin [ 151. As
a consequence, it is now more important that active
neurosyphilis be accurately diagnosed.
In this study we used relatively strict criteria for
diagnosing neurosyphilis in an attempt to minimize
classifying too many patients as having likely active
neurosyphilis. We required recent onset or progression of neurological signs consistent with acute syphilitic meningitis, general paresis, or meningovascular
syphilis to minimize the chances of classifying treated
neurosyphilis as active. Similarly, the signs of tabes
dorsalis were not included in our diagnostic criteria
because these findings can occur in burnt-out cases of
neurosyphilis [11] and in other diseases such as diabetes mellitus [6}. Since previous studies have reported
that the CSF of 30% of patients with well-diagnosed
general paresis either has a normal protein content or
is acellular [16], we required the CSF to show either a
pleocytosis or elevated protein. Finally, patients had to
have a reactive CSF-FTA-ABS test and no other recognized cause for the neurological illness. As a consequence of using these strict criteria, it is possible that
Davis and Schmitt: CSF Tests for Neurosyphilis 51
Table I. Characteristics of Patients with Likely Active Neurosyphilis
CSF Analysis
Age
Case (yr)/
No. Sex
Previous Diagnosis
and Treatment
of Syphilis
Clinical Information
1
56IM
No
2
44/M
Yes
3
79lM
Yes
4
32/M
Yes
5
73/F
No
6
6 1/M Yes
7
46lM
No
8
42/F
No
9
78lF
No
10
7 7/M
No
11
73lM
No
12
J 7/F
No
13
85/M
No
14
6UM
Yes
15
2 7lM
Yes
Recent memory
loss and personality change
Recent personality
change and psychiatric symptoms
Recent memory
loss and confusion
Chorioretinitis, vertigo, hearing loss
that cleared with
penicillin
Recent memory
loss and personality change
Recent memory
loss and personality change
Recent hemiparesis
and diplopia
Recent dementia
and unknown
chronic encephalitis
Recent memory
loss, seizures, irregular pupils
Recent brainstem
and cortical infarcts, dementia,
COPD, pneumonia
Recent dementia
and brainstem
stroke
Recent seizures and
psychiatric symptoms
Recent memory
loss and personality change, old
stroke, irregular
pupils
Progressive optic
atrophy and dementia, brain atrophy by CT
Recent dementia,
psychosis, HIV +
WBC/mm3
S e w Analysis
Protein
RBC/mm3 (mgldl) ETA-ABS
VDRL
FTA-ABS
RPR
16
0
106
4+
1:16
4+
1:64
70
1
75
4+
1:16
4+
1:512
1
297
340
3+
1:2
4+
1:2
16
0
4+
1:2
4+
1:2,048
15
40
141
3+
NR
3+
1:4
2
109
48
2+
NR
4+
1:16
7
18
73
2+
NR
4+
1:2
100
1,000
30
2+
NR
4+
1:128
6
8
36
2+
NR
3+
NR
7
0
52
2+
NR
4+
1:2
4
3
55
4+
NR
4+
1:2
3
1
58
2+
NR
4+
1:512
2
0
64
2+
NR
3+
ND
6
3
38
2+
NR
4+
1:512
2
0
68
1+
NR
4+
1:32
CSF = cerebrospinal fluid; WBC = white blood cells; RBC = red blood cells; ETA-ABS = fluorescent treponemal antibody absorption;
VDRL = Venereal Disease Research Laboratory; RPR = rapid plasma reagin; NR = nonreactive; N D = not done; COPD = chronlc
obstructive pulmonary disease; CT = computed tomography; HIV = human immunodeficiency virus.
52
Annals of Neurology Vol 25 No 1 January 1989
Table 2. Characteristics of Patients Unlikely Active Neumsyphilis
~
~
~~
CSF Analysis
Serum Analysis
Sex
Previous Diagnosis
andTreatmenr
Prorein
of Syphilis
CliNcalInfonnation \wBClmm3 R13C/mm3 (mgldl) JTA-ABS
VDRL
JTA-ABS
RPR
16
76/M
No
17
62/M
No
ia
771F
No
19
811F
No
20
21
6O/M Yes
87/M No
22
3 m
No
23
49m
No
24
731M
Yes
25
26
27
62/F Yes
3 1/M No
W M No
28
371M
Yes
29
57/M
Yes
30
31
17m
34M
Yes
No
32
33
47/M
54/F
No
Yes
34
76/F
Yes
35
37lM
Yes
36
37
79lF
6OlM
Yes
No
38
22lM
No
Age
Case
(yr)l
No.
1
Diabetes mellitus,
fever
1
Diabetes mellitus,
old stroke
Anoxic encephalop- 2
athy from cardiac
arrest
Parkinson’s disease,
0
headaches
4
Dementia, AIDS
0
Rend failure with
wmia
61
Neuromyelitis optics at autopsy
4
Seizures from alcohol and diazepam withdrawal
1
Normal neurological exam
1
Chronic anviety
0
Depression
0
Lacunar stroke at
autopsy
0
AIDS, normalneurological
exam
0
Recent dementia,
old hemiparesis
0
New seizures
4
Normal neurological exam
2
Low back pain
0
Irregular pupils,
sluggish pupil
reaction
UNK
Normal neurological exam
0
Normal neurological exam
0
Irregular pupils
0
Normal neurological exam, ear
disease
2
ARC and chronic
inflammatory
neuropathy
8
47
1+
NR
ND
ND
0
163
2+
NR
4+
1:2
0
45
2+
NR
ND
ND
0
24
1+
NR
3+
1:2
a2
690
75
39
1+
2+
NR
NR
4+
ND
1:16
ND
0
427
1+
NR
ND
ND
55
29
2+
NR
ND
NR
0
24
1+
NR
4+
i:a
380
2
1+
2+
1+
NR
NR
NR
3+
4+
3+
1:a
0
30
26
22
15
31
3+
NR
4+
1:256
165
30
2+
NR
1+
1:2
0
105
27
31
2+
2+
NR
NR
4+
4+
1:32
1:32
0
57
27
3+
6
2+
NR
NR
4+
3+
1:4
UNK
UNK
I+
NR
3+
1:32
1
35
2+
NR
3+
1:4
0
38
6
41
2+
1+
NR
NR
3+
3+
1:8
0
74
2+
NR
4+
1:32
1:16
1:16
1:64
1:4
AIDS = acquired immunodeficiency syndrome; UNK = unknown; ARC = AIDS-related complex. For other abbreviations see foomore to
Table 1.
Davis and Schmitt: CSF Tests for Neurosyphilis
53
some patients classified as unlikely to have active
neurosyphilis had, in fact, active disease. Nevertheless,
we identified 15 patients who were classiffied as having
likely active neurosyphilis. Only four (27%) had a
positive CSF-VDRL test. Thus, while the specificity of
this CSF-VDRL test in detecting active disease was
high, the sensitivity was quite low. On the other hand,
the CSF-FTA-ABS test appeared to be too sensitive
and had the potential to overdiagnose active disease.
Some generalizations can be made about the usefdness of both tests. First, a nonreactive serum FTAABS test rules out any syphilitic infection including
neurosyphilis [l2}. Experience by the CDC 171 and
others f 171 has shown that biological false-negative
tests are rare except possibly in individuals infected
with human immunodeficiency virus 118). Second, a
reactive serum FTA-ABS test with a nonreactive CSFFTA-ABS test makes a diagnosis of active neurosyphilis unlikely {8, 191. While this generalization has
not been universally accepted by those who diagnose
neurosyphilis on clinical grounds 14, 51, there have
been only rare reports of patients with a reactive CSFVDRL test and a nonreactive CSF-FTA-ABS test [5}
and no reports of T . pallidurn being isolated from CSF
or brain from a patient with late syphilis and a nonreactive CSF-FTA-ABS test. Third, a reactive CSF-VDRL
test makes a diagnosis of neurosyphilis likely {l].
However, this does not hold true for testing CSF with
the rapid plasma reagin test, in which false-positive
results are frequent 1201. While the serum VDRL test
may give false-positive reactions in a variety of diseases
t211, these diseases rarely result in a reactive CSFVDIU test [20, 22). However, the CSF-VDRL test
may be reactive in asymptomatic neurosyphilis [S, 231
and for a long time after adequate treatment of
neurosyphilis [24]. Therefore, a reactive test must be
combined with the presence of recent or progressive
neurological signs consistent with neurosyphilis and an
abnormal CSF when diagnosing active neurosphilis
1251. Unfortunately, as seen from this study, many
cases of likely active neurosyphilis lack a reactive CSFVDRL test. Fourth, a reactive CSF-FTA-ABS test can
indicate several things: active neurosyphilis, asymptomatic neurosyphilis, treated neurosyphilis, and a
false-positive reaction. Like the serum FTA-ABS test,
once the CSF-FTA-ABS becomes reactive, the reactivity persists for many years, even after adequate treatment has been given [26}. False-positive tests occur
when serum treponemal antibody leaks into the CSF
through a breakdown of the blood-brain-CSF barrier
or when a small amount of blood containing antibody
contaminates the CSF at the time of the lumbar puncture [lo}.
Three CSF tests may be of help in distinguishing a
truly positive CSF-FI'A-ABS test from a false-positive
54 Annals of Neurology Vol 25 NO 1 January 1989
test. The presence of increased intrathecal synthesis of
immunoglobulin G (IgG) (IgG synthesis rate increased
f271 or elevated IgG index 1281) or CSF oligoclonal
bands makes neurosyphilis more likely. Intrathecal
synthesis of IgG antibody to T . pallidurn occurs in
neurosyphilis, and oligoclonal bands in CSF have been
recognized in at least 75% of cases [26, 29, 307. The
extent of disease activity cannot be determined from
these two tests, as results may be abnormal in asymptomatic neurosyphilis 1231 and in treated neurosyphilis
[261. A recent study, however, reported elevated CSF
IgM levels in 6 of 6 patients with symptomatic neurosyphilis, but in none of 9 patients with asymptomatic
neurosyphilis and in only 11 of 498 patients with other
neurological diseases [3 11. Further experience with
this test will be needed to determine if it can reliably
distinguish active from asymptomatic neurosyphilis.
This work was supported by the Research Service of the Veterans
Administration.
For their help and advice in preparing this manuscript, we wish to
thank Drs E. Garland, T. Merlin, H. Raroque, R. Steece, and E.Umland; Ms E. Smith, Albuquerque, NM, and Dr S. Larsen, Centers for
Disease Control, Atlanta, GA.
References
1. Jaffe HW, Kabins SA. Examination of cerebrospinal fluid in
patients with syphilis. Rev Infect Dis 1982;4(suppl):842-847
2. Ch'ien L, Hathaway BM, Israel CW. Seronegative dementia
paralytica: report of a case. J Neurol Neurosurg Psychiatry
1970;33:376-3ao
3. Hooshmand H, Escobar MR, Kopf SW. Neurosyphilis. A study
of 241 patients. JAMA 1972;219:726-729
4. Wdkinson AE. Fluorescent treponemal antibody tests on cerebrospinal fluid. Br J Vener Dis 1973;49:346-349
5. Escobar MR, Dalton HP, Allison MJ. Fluorescent antibody tests
for syphilis using cerebrospinal fluid: clinical correlation in 150
cases. Am J Clin Pathol 1970;53:886-890
6. Swart2 M. Neurosyphilis. In: Holmes KK, Mardh PA, Sparling
PF, Wiesner PJ, eds. Sexually transmitted diseases. New York
McGraw-Hill, 1984:318-334
7. Deacon WE, Lucas JB, Price EV. Fluorescent trepunemal antibody-absorption (ETA-ABS) test for syphilis. JAMA 1966;
i9a:624-62a
8. Jaffe HW, Larsen SA, Peters M, et al. Tests for treponemal
antibody in CSF. Arch Intern Med 1978;138:252-255
9. Centers for Disease Control. Criteria and techniques for the
diagnosis of early syphilis. Atlanta: US Government Printing
Office, 1979. (DHEW publication no. 98-376)
10. Davis LE, Sperry S. The CSF-ETA test and the significance of
blood contamination. Ann Neurol 1979;6:68-69
11. Merritt HH, Adams RD, Solomon HC. Neurosyphilis. New
York Oxford U Pr, 1946
12. Simon RP. Neurosyphilis. Arch Neurol 1985;42:606-613
13. Centers for Disease Control. Manual of tests for syphilis1969. Atlanta: US Government Printing Office, 1969:1-81
(DHEW publication no. 41 1)
14. Greene BM, Miller NR, Bynum TE. Failure of penicillin G
benzathine in the treatment of neurosyphilis. Arch Intern Med
1980;140:1117-1118
15. Centers for Disease Control. 1985 STD treatment guidelines.
MMWR 1985;34:896
16. Dewhurst K. The composition of the cerebro-spinal fluid in the
neurosyphilitic psychoses. Acta Neurol Scand 1969;45:119123
17. Tuffanelli DL, Wuepper KD, Bradford LL, et al. Fluorescent
treponemal-antibody absorption tests. N En& J Med 1967;
276:258-261
18. Hicks CB, Benson PM, Lupton GP, Tramont EC. Seronegative
secondary syphilis in a patient infected with the human immunodeficiency virus (HIV) Kaposi sarcoma. Ann Intern Med
1987;107:492-495
19. McGeeney T, Yount F, Hinthorn DR, et al. Utility of the FTAAbs test of cerebrospinal fluid in the diagnosis of neurosyphilis.
Am Vener Dis Assoc 1979;6:195-198
20. Larsen SA, Hambie EA, Wobig GH, Kennedy EJ. Cerebrospinal fluid serologic test for syphilis: treponemal and nontreponemal tests. In: Morisset R,Kurstak E, eds. Advances in s e d y
transmitted diseases. Utrecht: VNU Science Press, 1987:157162
21. Catterall RD. Presidential address to the M.S.S.V.D. Systemic
disease and the biological false positive reaction. Br J Vener Dis
1972;48:1-12
22. Madiedo G, Ho KC, Walsh P. False-positive VDRL and FTA in
cerebrospinal fluid. JAMA 1980;244:688-689
23. Lowhagen GB, Anderson M, Blomstrand C , Roupe G. Central
nervous system involvement in early syphilis: Part I. Intrathecal
immunoglobulin production. Acta Derm Venereol (Stockh)
1983;63:409-4 17
24. Wdner E, Brody JA. Prognosis of general paresis after treatment. Lancet 1968;2:1370-1371
25. Dattner B, Thomas EW, De Mello L. Criteria for the management of neurosyphilis. Am J Med 1951;10463-467
26. van Eijk RVW, Wolters EC, Tutuarima JA, et al. Effect of early
and late syphilis on central nervous system: cerebrospinal fluid
changes and neurological deficit. Genitourin Med 1987;63:7782
27. Tourtellolte WW, Potvin AR, Fleming JO, et al. Multiple sclerosis: measurement and validation of central nervous system
IgG synthesis rate. Neurology 1980;30:240-244
28. Tebbling G, Link H, Ohman S . Principles of albumin and IgG
analysis in neurological disorders: I. Establishment of reference
values. Scand J Clin Lab Invest 1977;37:385-390
29. Vandal F, Vandvik B, Michaelsen TE, et al. Neurosyphhs: intrathecal synthesis of ohgoclonal antibodies to Treponaa pallzdum. Ann Neurol 1982;11:35-40
30. Muller F, Moskophidis M. Estimation of the local production of
antibodies to Treponema pallidum in the central nervous system
of patients with neurosyphilis. Br J Vener Dis 1983;59:80-84
31. Lee JB, Farshy CE, Hunter EF, et al. Detection of immunoglobulin M in cerebrospinal fluid from syphilis patients by enzyme-linked immunosorbent assay. J Clin Microbiol 1986;24:
736-740
Davis and Schmitt: CSF Tests for Neurosyphilis
55
Документ
Категория
Без категории
Просмотров
12
Размер файла
492 Кб
Теги
test, clinical, neurosyphilis, significance, fluid, cerebrospinal
1/--страниц
Пожаловаться на содержимое документа