вход по аккаунту


Evidence of an immunopathogenic basis for central nervous system disease in primary Sjgren's syndrome.

код для вставкиСкачать
The pathogenesis of central nervous system complications in primary Sjogren’s syndrome (CNS-SS) is
unknown. In order to determine whether patients with
active CNS-SS have cerebrospinal fluid (CSF) abnormalities indicative of CNS inflammation, CSF analyses
from 30 patients with active CNS-SS (SSA) were contrasted with those from 20 SS patients without CNS
involvement (SSI) and 20 patients with systemic lupus
erythematosus and active CNS disease (SLEA).Elevations of total protein concentration, IgG concentration,
IgG to total protein ratio, and IgG index were observed
in patients with SSA,but not in those with SSI. Agarose
gel electrophoresis results were abnormal, with 1 or
more bands, in 25 of 29 SSApatients (86%), but in only
3 of 18 SSI patients (17%). Similar, but less striking,
CSF abnormalities were seen in a minority of SLEA
patients. Fifteen SSApatients (50%) had transient, mildto-moderate CSF pleocytosis, while only 1 SSI patient
and 2 SLEA patients had similar findings. Cytologic
findings were abnormal in 18 SSApatients (60%); these
included atypical mononuclear cells, lymphoblastoid
From the Johns Hopkins University School of Medicine,
Baltimore, Maryland.
Supported in part by NIH grants HL-30748 and NS-00632
and by the Kroc Foundation.
Elaine L. Alexander, MD, PhD: Department of Medicine,
Clinical Immunology Division, Johns Hopkins University School of
Medicine; Jane E. Lijewski, BS, MT, ASCP: Department of Laboratory Medicine, Johns Hopkins University School of Medicine;
Myles S. Jerdan, MD: Department of Dermatology, Johns Hopkins
University School of Medicine; Garrett E. Alexander, MD, PhD:
Department of Neurology, Johns Hopkins Medical Institutions.
Address reprint requests to Elaine L. Alexander, MD, PhD,
The Johns Hopkins Medical Institutions, Department of Medicine,
Division of Clinical Immunology, Good Samaritan Hospital, 5601
Loch Raven Boulevard, Baltimore, MD 21239.
Submitted for publication August 16, 1985; accepted in
revised form March 20. 1986.
Arthritis Rheumatism, Vol. 29, No. 10 (October 1986)
cells, and plasma cells. The presence of immunocompetent cells and evidence for the intrathecal synthesis of IgG within the CSF of SSA,but not SSI, patients
provide diagnostic parameters which are indicative of
active disease and which can be monitored serially
during therapy.
Sjogren’s syndrome (SS) is a common rheumatic disorder (1); among rheumatic diseases, only
rheumatoid arthritis (RA) exceeds SS in prevalence
(2). Recent epidemiologic studies indicate that SS may
be even more common than previously estimated
(3,4). Central nervous system (CNS) complications of
primary SS (CNS-SS) are becoming recognized increasingly. A spectrum of CNS manifestations has
been observed, including focal neurologic deficits,
psychiatric disturbances, cognitive dysfunction, seizure of movement disorders, recurrent aseptic
meningoencephalitis (AME), and acute and chronic
myelopathy (5-9).
The pathogenesis of these diverse CNS manifestations in SS remains unknown. Indirect evidence
suggests that vascular inflammation may be involved
in some cases. By biopsy, we have documented concomitant peripheral inflammatory vascular disease
(IVD) (involving skin, muscle, or nerve) in up to 75%
of SS patients with active CNS disease (6,lO). Furthermore, in a group of 50 SS patients with biopsydocumented peripheral IVD, we have observed a high
proportion (66%) with active CNS disease (1 1).
There is little direct histopathologic documentation of the presence of CNS vascular inflammation in
CNS-SS. Our preliminary observations on postmortem and biopsy specimens of brain tissue from patients
with CNS-SS indicate that there may be a cerebral
vasculopathy (Alexander et al: unpublished observations), and we have described an SS patient with
necrotizing spinal arteritis (7). Other investigators
have documented the presence of IVD in SS patients
with active nervous system disease (1,12,13).
It is difficult to obtain objective documentation
of active CNS disease in patients with connective
tissue diseases (CTD). CSF analysis provides a potential means of assessing the presence of intrathecal
inflammation ante mortem. An elevated IgG index has
been established as a reliable indicator of the intrathecal production of IgG (14,15), particularly in the presence of an intact blood-brain barrier (16). The use of
agarose gel electrophoresis has led to the description
of specific qualitative abnormalities of immunoglobulin, and these consist of restricted bands of protein,
each representing immunoglobulin secreted by individual clones of B cells (17,18). Such CSF immunoglobulin abnormalities have been observed in
several autoimmune neurologic diseases (19), and
most recently in CNS disease associated with systemic
lupus erythematosus (SLE) (20,21), in which immunemediated pathogenic mechanisms are implicated.
Limited observations suggest that CSF abnormalities may reflect CNS inflammation in Sjogren’s
syndrome. Attwood and Poser (22) described an SS
patient with multiple cranial nerve palsies and peripheral neuropathy, whose CSF showed an increased
protein concentration, pleocytosis, and an abnormal
colloidal gold curve. The recent description of recurrent AME in primary SS patients (8) suggested that the
leptomeninges also might be infiltrated by inflammatory cells; this was subsequently confirmed in postmortem studies (23). Evaluation of the CSF from SS
patients with AME showed an increased IgG index
and/or an abnormal agarose gel electrophoretic immunoglobulin pattern, in addition to pleocytosis with
lymphoblastoid and plasma cells (8).
To gain further information regarding the immunopathogenesis of CNS-SS, we examined the CSF
IgG index, CSF agarose gel electrophoretic immunoglobulin pattern, and CSF cytology of 30 patients
with primary SS who had active CNS disease (SSA)
unattributable to other causes. We contrasted these
CSF findings with those of a group of 20 patients with
primary SS without CNS disease (SSI) and with those
of a group of 20 patients with SLE and active CNS
disease (SLEA). We present evidence of the presence
of reactive lymphoid cells and the intrathecal synthesis
of IgG (elevated IgG indices and 1 or more clonal
bands) in the CSF of SSA,but not SSI, patients.
Patient population. Thirty patients with SSA,20 with
SSI, and 20 with SLEA, all of whom were observed at the
Johns Hopkins Medical Institutions, were analyzed retrospectively. Of the 50 patients with primary SS, none met the
criteria for another CTD (i.e., SLE [241, RA [251, or systemic sclerosis [26]).The clinical features of 4 SS patients
with AME (8) and 2 patients with other manifestations of
CNS disease (6) have been described previously. All SLE
patients met 4 or more of the American Rheumatism Association revised criteria for SLE (24).
Patients were selected for study only if they were not
treated with immunosuppressive agents at the initially studied time point in their medical history, and if at least 2 of the
following 3 CSF analyses had been performed: IgG index,
agarose gel electrophoresis, and cytologic examination. All
SS patients were evaluated by at least 1 neurologist in
addition to 1 of the authors, and the deficits reported reflect
this evaluation. The categorization of patients with active
versus inactive nervous system disease was made independently of the investigator’s knowledge of CSF findings. The
CNS disease in SSAand SLEApatients was considered to be
secondary to their underlying CTD; CNS disorders could not
be accounted for by associated disease processes such as
hypertension, diabetes mellitus, arteriosclerosis, infection,
embolization, or oral contraceptives.
The 20 patients with SS1 had CSF analyses performed during evaluation for clinical manifestations not
believed to represent active CNS disease secondary to CTD
(e.g., headache, low back pain, or cervical or lumbar
radiculopathy secondary to degenerative osteoarthritis).
Four SSIpatients had a remote history of CNS disease which
had resolved or stabilized without therapy.
Sicca complex. All patients with primary SS had sicca
symptoms: xerophthalmia, xerostomia, or recurrent or
chronic salivary gland enlargement. The sicca complex was
documented as outlined previously (10). Xerophthalmia was
evaluated by Schirmer’s and rose bengal dye tests. Decreased lacrimation was documented by the Schirmer’s test
(27); the result was considered abnormal if <5 mm of filter
paper was moistened after 5 minutes. Characteristic rose
bengal staining (Holm type A or B) (28) or a positive result
on slit-lamp examination was required for the diagnosis of
keratoconjunctivitis sicca (KCS). Xerostomia was considered present when, in symptomatic patients, desiccated
mucous membranes and loss of the sublingual salivary pool
were observed. Salivary gland enlargement was either recurrent or chronic. Minor salivary gland biopsies were scored as
described previously, and biopsies graded 3 or 4 by the
criteria of Greenspan et a1 (29) were considered positive.
CSF analysis. CSF was obtained by lumbar puncture
in the course of routine clinical evaluation. Total white blood
cell (WBC) and differential cell counts were performed
immediately. Cytologic examination was performed on
5.0-1111 specimens of CSF which had been gently aspirated by
filtration onto Millipore filters (Millipore, Bedford, MA) or
prepared by cytocentrifugation with the Shandon cytocentrifuge (Shandon Southern Instrument, Inc., Sewickley,
PA) by standard methods. The specimens were stained with
Giemsa stain.
All CSF samples for noncytologic studies were either
evaluated immediately or frozen immediately and stored at
-70°C until analysis. No specimen underwent freezethawing more than once prior to testing. CSF and simultaneous peripheral blood glucose, as well as CSF protein and
IgG determinations, were made by standard methods.
Abnormalities in level of CSF protein, CSF IgG
concentration, and CSF IgG to total protein ratio were
defined as >45 mg/dl, >6 mg/dl, and >14%, respectively.
CSF and serum albumin and IgG concentrations were measured simultaneously by radial electroimmunodiffusion (30).
The CSF IgG index was calculated by the formula: (CSF
IgGlCSF albumin)/(serum IgGherum albumin) (16,17). CSF
IgG indices 20.7 were considered abnormal (16) and indicative of endogenous production of IgG within the CNS. The
CSF albumin x 103:serum albumin ratio (Q albumin), a
measure of blood-brain barrier integrity, was considered
abnormal when it was >9 (16,17). The degree of blood-brain
barrier impairment was graded according to the method of
Schliep and Felgenhauer (31): P-14.3 = slight; 14.3-33.3 =
moderate; 33.3-100 = severe; >lo0 = total breakdown.
Agarose gel electrophoresis was performed by standard techniques (32) using the Panagel electrophoresis system (Worthington Diagnostics, Freehold, NJ). The CSF was
concentrated eightyfold and the serum was diluted 1:4.
Sample volumes of 10 ~1 were applied to the gel. The
electrode buffer was barbituric acid/sodium barbituate, pH
8.6, 0.05 moles/liter, containing 0.576 gm/liter calcium lactate. The Panagel migration unit delivered a constant voltage
of 200 DC. Electrophoresis was performed for 45 minutes.
The gels were stained with Coomasie blue (0.2% Kodak
Coomassie Brilliant Blue R-250, EKC 14013 [Scientific Products, McGaw Park, IL] in methanoVacetic acid/water). Immunofixation was performed with picric acid fixative
(DR0047611; Worthington).
Control samples consisted of normal CSF, CSF from
a patient with classic multiple sclerosis (MS) (which produced oligoclonal bands), and serum from a patient with
multiple myeloma (which consistently produced 2 discrete
bands in the gamma globulin region (Figure 1). Only thin
discrete bands, rather than diffuse broad bands indicative of
polyclonal immunoglobulin, were scored as abnormal. The
results represent the official clinical laboratory interpretation. Myelin basic protein determinations were performed by
radioimmunoassay (33), and levels >5 mg/dl were considered abnormal (33). Reference values for CSF IgG determinations, IgG indices, albumin quotients, oligoclonal banding
patterns, and myelin basic protein determinations were
verified with control sera from normal individuals and from
patients with classic MS.
Microbiologic studies. The CSF cultures from each
patient were negative for aerobic and anaerobic bacteria,
mycobacteria, viruses, and fungi. Cryptococcal antigen and
antibody determinations, as well as VDRL testing, were
performed on specimens from all patients with active CNS
disease, and all were negative.
Statistical methods. Clinical evaluation and CSF analyses were made independently, and then the findings were
correlated. Differences between clinical subgroups, and the
correlation of lymphoid cells with IgG indices and
oligoclonal bands were analyzed using Fisher’s 2-tailed
Figure 1. Agarose gel electrophoresis. Lanes 1-3, Cerebrospinal
fluid (CSF) from patients with Sjogren’s syndrome, showing: 1 band
(lane I), 2 bands (lane 2), and 3 bands (lane 3). Lanes 4-7, Controls.
Lane 4, Serum from patient in lane 3. Lane 5 , CSF from a normal
control. Lane 6, CSF from a patient with multiple sclerosis, showing
5 bands. Lane 7, Serum from a patient with multiple myeloma,
showing 2 bands.
exact test. Comparison of the frequency of occurrence of
events between groups was made using Yates’ corrected
chi-square analysis.
Demographic features and sicca complex. T h e
demographic and sicca features of the patient groups
are presented in Tables 1 and 2. E a c h of the 30 SSA
and 20 SSI patients, by selection, was symptomatic
with respect t o the sicca complex, and each had KCS,
a positive minor salivary gland biopsy, or both. In 18
(60%) of the SSA patients, Sjogren’s syndrome had
been diagnosed prior t o the development of CNS
disease, while in 12 patients (40%), the diagnosis was
established at the time of presentation with neurologic
complications. There were no differences in the
neurologic or CSF characteristics between the patients
with prior diagnosis of SS and those with concomitant
In all patients with primary SS, a second CTD
had been excluded. The SLEA patients were not
overtly symptomatic for the sicca complex, but had
not been evaluated for SS with detailed review of
Table I . Demographic features of patients with Sjogren’s
syndrome and active central nervous system involvement (SSA), SS
without nervous system involvement (SSI), and systemic lupus
erytheniatosus and active central nervous system involvement
(n = 30)
(n = 20)
(n = 20)
55 (27-75)
45 (12-80)
36 (19-56)
Age, mean (range)
symptoms, ophthalmologic examination, or minor salivary gland biopsy.
Neurologic manifestations. The diverse neurologic manifestations of patients with SSA are presented
in Table 3. Of the 30 SSA patients with CNS involvement, 23 (77%) had involvement of the brain and 11
(37%) had involvement of the spinal cord. Four patients had involvement of both. A total of 61 CNS
abnormalities occurred in the 30 SSA patients, and
two-thirds had multifocal CNS events. Disturbances
were recurrent over time in 26 (87%). Concomitant
peripheral nervous system disease was seen in
12 SSA patients (40%). This included peripheral neuropathy (8 patients), carpal tunnel syndrome (5 patients), and cranial neuropathy (3 patients).
The spectrum of nervous system manifestations
demonstrated by SLEA patients was not significantly
different from that of SSA patients (data not shown).
Nineteen SLEA patients had involvement of the brain
(focal in 17 and diffuse in 15), and 4 had involvement of
the spinal cord. Peripheral nervous system disease
was present in 4.
CSF findings. The CSF protein and IgG profiles
for the 3 patient groups are presented in Table 4.
Significant elevations in the CSF parameters of total
protein, IgG concentration, 1gG:total protein ratio,
and IgG index were found in the SSA group, when
compared with those in the SSI group. Moreover, a
significantly higher proportion of the SSA patients had
Table 2. Sicca features in patients with Sjogren’s syndrome
Positive result on Schirmer’s test
Keratoconjunctivitis sicca
Salivary gland enlargement
Minor salivary gland biopsy
* See Table I for definitions.
t Twent y-seven patients biopsied.
Table 3. Neurologic manifestations in patients with Sjogren’s
syndrome and active central nervous system involvement (SSA)
Neuroanatomic site
= 30)
= 61)
Seizure disorder
Movement disorder
Cognitive dysfunction
Aseptic meningoencephalitis
Spinal cord
Chronic progressive
Acute transverse myelopathy
Neurogenic bladder
elevated CSF total protein, IgG concentration, and
IgG index, in comparison with the SSI patients. In
contrast, the mean values for the IgG-related parameters (IgG concentration, IgG:total protein ratio, and
CSF IgG index) of the SSA group did not differ
significantly from those of the SLEAgroup. The only
significant difference between the 2 groups with active
CNS disease was a higher mean total protein value
among the SSA patients. The differences between the
SSI and SLEA groups were not significant for any of
these parameters.
The mean CSF:serum albumin ratios, which are
indicative of gross alterations in blood-brain barrier
permeability, were within normal limits for all groups
tested. Only 1 patient, with SSA, had a slightly abnormal value (14.0).
The results of CSF agarose gel electrophoresis
are presented in Table 5. Representative electrophoretic patterns are illustrated in Figure 1. A significantly
higher proportion of SSA patients had 1 or more
oligoclonal bands, which migrated in the region of IgG,
than did either the SSr patients or the SLEApatients.
The SSI and SLEAgroups did not differ in the proportion of patients with bands. The majority of the SSA
patients with abnormal CSF electrophoretic patterns
had a single discrete band, while approximately onethird had 2 bands. Two patients had patterns of
multiple bands which were indistinguishable from
those seen in MS. Each of the SSI patients with
oligoclonal bands had a remote history of nervous
system dysfunction which had stabilized, and none
had received treatment.
Paired serum and CSF samples from 14 patients
were analyzed simultaneously, by agarose gel electro-
Table 4.
Evidence of intrathecal IgG synthesis in patients with SSA, SSI, and SLEA*
Mean f SD
Total protein
54.8 f 5.6t
7.5 2 1.45
CSF IgG: total
13.3 f 1.887
CSF:serum albumin 6.5 f 0.49
CSF IgG index
1.0 2 0.16#
(n = 20)
(n = 20)
(n = 30)
No. (%)
No. (%)
abnormal Mean f SD abnormal
No. (%)
f 1.6
f 0.48
f 1.17
1 (5)
2 (10)
2 (10)
39.9 f 2.90
5.4 f 0.80
13.1 f 1.15
6 (30)
5 (25)
5 (25)
1 (3)
6.3 f 0.51
21 (70)** 0.49 f 0.49
0 (0)
4 (20)
0 (0)
8 (40)
16 (53)$
13 (43)$
11 (37)
* Abnormal values defined as follows: total protein, >45 mg/dl; cerebrospinal fluid (CSF) IgG, >6 mg/
dl; CSF IgG: total protein, > 14%; CSF albumin ( x lo3):serum albumin, >9; CSF IgG index, 20.7. See
Table 1 for definitions.
t P = 0.001 versus SSI group and P = 0.02 versus SLEAgroup, by Fisher’s 2-tailed exact test.
$ P 5 0.05 versus SSI group, by Yates’ chi-square analysis.
5 P = 0.005 versus SSr group, by Fisher’s 2-tailed exact test.
ll P = 0.04 versus SSI group, by Fisher’s 2-tailed exact test.
# P = 0.004 versus SSI group, by Fisher’s 2-tailed exact test.
** P 5 0.005 versus SS, group, by Yates’ chi-square analysis.
phoresis. Thirteen patients with 1 or more CSF bands
demonstrated no comparable discrete band(s) within
the serum. One patient with 2 CSF bands had both a
broad polyclonal band and a single discrete band in the
serum. The electrophoretic mobility of the serum
band, however, differed from that of the CSF bands.
This patient had a polyclonal gammopathy with a
serum globulin level of 8.0 mg/dl and high titers of
antibodies to Ro (SS-A).
All 30 SSApatients had an elevated IgG index,
1 or more oligoclonal bands, or both, in contrast to 7
(35%) of the SSI patients (P5 0.0005) and 11 (55%) of
the SLEA patients (P 5 0.0005).
Myelin basic protein (MBP) determinations
were carried out on all SSA patients, but CSF MBP
was elevated strikingly (36 mg/dl) in only 1, in association with the sequential development of multifocal
neurologic deficits (right hemiparesis, AME, and
transverse myelopathy). The patient was treated with
corticosteroids which led to resolution of the CNS
abnormalities and a return to normal of the MBP.
Subsequently, she had a recurrent right hemiparesis
and 2 episodes of optic neuritis, not associated with
changes in MBP. Two other SSA patients, 1 with
transverse myelopathy and the other with severe progressive cognitive dysfunction and a movement disorder, had mild elevations of CSF MBP (7.0 mg/dl and
9.7 mg/dl, respectively). MBP determinations were not
performed routinely in SSI or SLEApatients, but were
normal when studied.
The cellular characteristics of the CSF are
presented in Table 6. The proportion of SSA patients
with a CSF pleocytosis (>5.0 WBC/mm3) was significantly greater than the proportions observed among
SSI and SLEA patients. There was no difference in
CSF pleocytosis between the SSI and SLEA groups.
Cytologic examination of the CSF (Figure 2)
showed abnormalities in 18 SSA patients (60%) and
demonstrated small lymphocytes, plasma cells, and
lymphoblastoid cells, as well as large, often atypicalappearing mononuclear cells with convoluted nuclei,
prominent nucleoli, and relatively sparse cytoplasm.
Neither the 19 SSI nor the 17 SLEA patients examined
had similar cells on cytologic examination of the CSF.
Serial analyses of CSF IgG indices and oligclonal
bands. Fourteen SSA patients had more than 1 CSF
analysis performed; the results are presented in Table
7. Of 10 SSA patients treated with corticosteroids, all
had a decrease, by at least 0.2, in the IgG index. Five
of 9 of these patients also had a decrease in the number
Table 5. Detection of oligoclonal bands on agarose gel electrophoresis in patients with SSA, SSI, and SLEA*
No. of positive
oligoclonal bands
(n = 29)
(n = 18)
(n = 18)
Total with positive
bands. no. (%)
25 (86H
3 (17)
6 (33)
* See Table 1 for definitions.
t P 5 0.0005 versus each of the other 2 groups, by Yates’ chi-square
Table 6. Presence of lymphoid cells within the cerebrospinal fluid in patients with SSA, SS,, and
No. (%)
White blood cell count
46 f 30.5
15/30 (50)t
18/30 (60)§
f 0.3
No. (%)
No. (%)
Mean 2
1/20 (5)
0119 (0)
4.5 f 3.7
2/20 (10)
0117 (0)
* See Table 1 for definitions.
t P 5 0.005 versus SS, group and P 5 0.01 versus SLEA group, by Yates’ chi-square analysis.
S Lymphocytes, lymphoblastoid cells, and plasma cells.
5 P 5 O.OOO5 versus each of the other 2 groups, by Yates’ chi-square analysis.
of oligoclonal bands. In these patients and in the
others, the intensity of the band(s) decreased over
time. One such patient initially had 5 oligoclonal
bands, and on subsequent analyses the number of
bands decreased to 4 and then to 3 and subsequently to
2. Of the 4 SSApatients who were not treated, the IgG
index changed in 2, but the oligoclonal banding patterns remained unchanged. CSF pleocytosis with abnormal cells on cytologic examination also resolved
with therapy. Similar data were not available for the
m~al l umber Of SLEA Patients with b ~ e a s e d
indices and oligoclonal banding.
The present study represents the first analysis
of CSF from patients with primary SS and active CNS
disease. The data provide evidence of the presence of
immunologically competent cells within the CNS and
of the intrathecal synthesis of IgG. Although these
CSF findings are specific for a pathologic process,
rather than for a particular disease entity, their presence does suggest an immunopathogenic basis for
We found that the CSF IgG index was elevated
in 70% of SSA patients, while 40% of the SLEA
patients had abnormal values. Two studies of SLE
patients have examined the CSF IgG index (20,21).
Using the CSF IgG index reference value (20.7)
employed in this study, Seibold et a1 (20) demonstrated
elevated CSF IgG indices in 8 (40%) of 20 SLEA
patients, while Winfield et a1 (21) found them to be
elevated in 4 (21%) of 19 SLEA patients. These observations suggest that an elevated CSF IgG index, while
FigurP 2. Cytologic findings in the cerebrospinal fluid of Sjogren’s
synduome (SS) patients with central nervous system involvement.
A, Nornial small lymphocyte, demonstrating dense nuclear chromatin and scant cytoplasm. B, Activated lymphocyte, which is slightly
larged than a normal lymphocyte. The nucleus is convoluted, with a
promhent nucleolus and homogeneous cytoplasm. C, Plasma cell
with eccentric round-to-oval nucleus and pale cytoplasm. Blocks of
chromatin are arranged in a “clock-face’’ pattern. D, Atypical
monanudear cell with a large convoluted nucleus with irregular
aggregates of chromatin. The cytoplasm is relatively diminished and
homageneous. Similar cells have been observed within cerebral
blood vessels and meninges of SS patients. (Giemsa stained, original
magnification X 1,900.)
Table 7. Effect of corticosteroid therapy on cerebrospinal fluid
IgG index and oligoclonal bands in patients with Sjogren’s syndrome
and active central nervous system disease
IgG index
Corticosteroidtreated (n = 10)
No corticosteroids (n = 4)
Oligoclonal bands
occurring in both SSA and SLEA, may be more common in CNS-SS than in CNS-SLE. It has not been
determined, however, either in this study or in previous studies (20,21), whether the SLE patients with
abnormal indices may actually have had primary SS
with CNS involvement, or whether a subgroup of SLE
patients had associated SS.
CSF electrophoretic patterns in CNS-SLE have
been examined in 2 studies (20,21). Using agarose gel
electrophoresis, broad (rather than discrete, as in the
present study) immunoglobulin-containing bands were
observed by Seibold et a1 (20), in 33 of 34 patients with
SLEA (97%), but in only 1 of 12 SLEI patients (8%).
Winfield et a1 (21) described oligoclonal IgG banding
patterns detected by isoelectrophoretic focusing in 8
of 19 SLEApatients (42%). Neither of these studies,
however, reported the number of bands present in the
CSF samples from different individuals. In our study,
in which oligoclonal bands were present in the CSF of
almost 90% of SSApatients, the presence of either 1 or
2 oligoclonal bands was most common, but 2 patients
in our series, as well as other, previously described SS
patients with CNS disease (34), had multiple bands.
Although the morphology of the latter banding pattern
is indistinguishable from that seen in MS (32,34),
oligoclonal bands in CNS-SS tend to decrease in
number and intensity following corticosteroid therapy,
while in MS they persist (35).
Although a mild CSF pleocytosis has been
observed in a minority of CNS-SLE patients (36-42),
particularly in the setting of AME and in transverse
myelopathy, descriptions of cytologic abnormalities in
CNS-SLE are rare (38,39,42). Reactive mononuclear
and lymphoid cells which morphologically resemble
those seen in the CSF of our SSA patients have not
been described in CNS-SLE, although in a case report
of a patient with SLE and transient cortical blindness,
mononuclear phagocytes composed 55% of the total
number of CSF cells (43).
To our knowledge, cytologic examination of the
CSF from a series of patients with SS has not been
performed previously. A spectrum of reactive
lymphoid cells (including large mononuclear cells with
convoluted nuclei) was observed in patients with SSA,
but not in those with SSI or SLEA. These cells
resemble those we have described in the CSF of SS
patients with AME (8). Although reactive lymphoid
cells, the majority of which are T helper cells, also
have been observed in the CSF of patients with MS
(44,45), atypical mononuclear cells similar to those in
CNS-SS have not been described. Reactive lymphoid
cells may represent an inflammatory response associated with, but not specific for, SS. Nevertheless, their
absence in CNS-SLE may be an important differential
feature or, alternatively, may merely reflect the lower
total CSF WBC count (Table 6) in the SLEApatients in
this series.
All SSA patients with plasma cells or other
potentially immunocompetent lymphoid cells within
the CSF had an abnormal CSF IgG index, an abnormal
banding pattern on CSF agarose gel electrophoresis,
or both. This concordance suggests that such
lymphoid cells may be responsible for the intrathecal
production of IgG in SSA. Morphologically identical
cells within the leptomeninges (15) and within and
around cerebral blood vessels have been observed in
pathologic specimens from SS patients (Alexander et
al: unpublished observations).
There was evidence of mild impairment of the
blood-brain barrier permeability (abnormal CSF:
serum albumin ratio) in only 1 of 30 SSApatients and
in none of the SSI or SLEApatients in this series. This
result is similar to that obtained in a recent study of
CNS manifestations in SLE (20), in which only 1 of 20
patients (5%) was found to have a CSF:serum albumin
ratio indicative of abnormal blood-brain barrier function. These data, however, do not exclude focal alterations of blood-brain barrier permeability in individual
vessels, which would not be detected by this technique. These observations stand somewhat in contrast
to those of Winfield et a1 (21), who reported an abnormal CSF:serum albumin ratio in one-third (6 of 19)
of CNS-SLE patients.
Evidence which indicates a grossly intact
blood-brain barrier, associated with the presence of
elevated IgG indices and abnormal electrophoretic
banding patterns, is consistent with the intrathecal
synthesis of IgG, rather than passive diffusion of IgG
across the blood-brain barrier (16). (There is no evidence, in either healthy or diseased subjects, for the
active transport of immunoglobulin from blood to CSF
[14-16,18,19].) The potential specificity of such
intrathecally synthesized IgG is unknown, but such
antibodies could have direct effects on CNS function if
directed against neural antigens or neurotransmitters.
It is not known whether intrathecal synthesis of antibodies is a primary event or whether such antibodies
play a role in the pathogenesis of CNS-SS. Alternatively, intrathecal production of immunoglobulin in
CNS-SS may represent either nonspecific activation of
immunocompetent cells that have gained incidental
access to the CNS (e.g., as a result of primary vascular
inflammation) or specific, but secondary, synthesis of
antibodies directed against antigens exposed b y prior
CNS tissue damage.
T h e presence of elevated C S F IgG/albumin
indices, oligoclonal banding, or abnormal cytology
serves to distinguish the majority of SSApatients from
SSI patients. Data available from a subgroup of SSA
patients treated with corticosteroids, whose CSF was
analyzed serially, indicate that the CSF IgG albumin
index usually decreases and the number (or intensity)
of oligoclonal bands may decrease with therapy. CSF
pleocytosis, when present, also resolves with therapy.
These observations suggest that serial CSF analyses
assessing these 3 features may be of assistance in
establishing the presence of active CNS disease in SS
patients and, if the disease is present, in monitoring
disease activity and response to therapy. T h e absence
of these parameters, however, does not exclude t h e
presence of active CNS-SS. We do not yet know if
these abnormalities serve to distinguish SSA or
SS/SLE* overlap patients from SLEA patients. W e
recommend that the CSF of CTD patients with potentially active CNS disease be examined f o r these markers of CNS inflammation initially, at the time of
presentation, and serially.
The authors would like to express their appreciation
to Dr. Barry G. W. Arnason for helpful discussions; to
Dr. Sheldon Glusman, Dr. Donald Edlow, Therese B.
Datiles, and Roger Frye for their assistance in these studies;
and to Vicky Rogers for her professional preparation of the
1. Shearn MA: Sjogren’s syndrome, Major Problems in
Internal Medicine. Vol. 11. Edited by LH Smith. Philadelphia, WB Saunders, 1971, pp 12-14
2. Hochberg MC: Adult and juvenile rheumatoid arthritis:
current epidemiologic concepts. Epidemiol Rev 3:27-44,
3. Reveille JD, Hochberg MC, Bias WB, Arnett FC: Rheumatic diseases and autoantibodies in the old order
Amish: prevalence and HLA associations (abstract).
Arthritis Rheum (suppl) 26:S40, 1983
4. Strickland RW, Tesar JT, Berne BH, Hobbs BR,
Moeller DR, Richardson DD, Lewis DM, Welton RC:
The prevalence of Sjogren’s syndrome and associated
rheumatic diseases in an elderly population (abstract).
Arthritis Rheum (suppl) 27:S45, 1984
5. Alexander GE, Provost TT, Stevens MB, Alexander
EL: Sjogren’s syndrome: central nervous system manifestations. Neurology 31: 1391-1396, 1982
6. Alexander EL, Provost TT, Stevens MB, Alexander
GE: Neurologic complications of primary Sjogren’s syndrome. Medicine (Baltimore) 61:247-257, 1982
7. Alexander EL, Craft C, Dorsch C, Moser R, Provost
TT, Alexander GE: Necrotizing arteritis and spinal
subarachnoid hemorrhage in Sjogren’s syndrome. Ann
Neurol 11:632-635, 1982
8. Alexander EL, Alexander GE: Aseptic meningitis in
primary Sjogren’s syndrome. Neurology 33593-598,
9. Malinow KL, Molina R, Gordon B, Selnes OA, Provost
TT, Alexander EL: Neuropsychiatric dysfunction in
primary Sjogren’s syndrome. Ann Intern Med 103:344350, 1985
10. Alexander EL, Arnett FC, Provost TT, Stevens MB:
Sjogren’s syndrome: association of anti-Ro(SS-A) antibodies with vasculitis, hematologic abnormalities, and
serologic hyperreactivity. Ann Intern Med 98: 155-159,
11. Molina R, Provost TT, Alexander EL: Peripheral inflammatory vascular disease in Sjogren’s syndrome:
association with nervous system complications (abstract). Arthritis Rheum (suppl) 28:S36, 1985
12. Kaltreider HB, Talal N: The neuropathy of Sjogren’s
syndrome: trigeminal nerve involvement. Ann Intern
Med 70:751-762, 1969
13. Pittsley RA, Talal N: Neuromuscular complications of
Sjogren’s syndrome, Handbook of Clinical Neurology.
Vol. 139. Edited by PJ Vinken, GW Bruyn. Amsterdam,
ElseviedNorth Holland, 1980, pp 419433
14. Tibbling G, Link H , Ohman S: Principles of albumin and
IgG analyses in neurologic disorders. I. Establishment
of reference values. Scand J Clin Lab Invest 37:385-390,
15. Link H, Tibbling G: Principles of albumin and IgG
analyses in neurologic disorders. 11. Relation of the
concentration of the proteins in serum and cerebrospinal
fluid. Scand J Clin Lab Invest 37:391-396, 1977
16. Lefvert AK, Link H: IgG production within the central
nervous system: a critical review of proposed formulae.
Ann Neurol 17:13-20, 1985
17. Mattson DH, Roos RP, Arnason BGW: Comparison of
agar gel electrophoresis and isoelectric focusing in multiple sclerosis and subacute sclerosing panencephalitis.
Ann Neurol 9:34-41, 1985
18. Walsh MJ, Tourtellotte WW, Potvin AR, Potvin JH: The
cerebrospinal fluid in multiple sclerosis, Multiple Sclerosis: Pathology, Diagnosis and Management. Edited by
JF Hallpike, CWM Adams, WW Tourtellotte. London,
Chapman & Hall, 1983, pp 275-358
19. Maurice P, Ma BI, Tourtellotte WW: In situ central
nervous system IgG synthesis, cerebrospinal fluid
oligoclonal IgG, and blood-brain barrier in multiple
sclerosis and other neurologic disease. Neurology
27~372-373, 1977
20. Seibold JR, Buckingham RB, Medsger TA, Kelly RH:
Cerebrospinal fluid immune complexes in systemic
lupus erythematosus involving the central nervous system. Semin Arthritis Rheum 12:68-75, 1982
21. Winfield JB, Shaw M, Silverman LM, Eisenberg RA,
Wilson HA, Koffler D: Intrathecal IgG synthesis and
blood-brain barrier impairment in patients with systemic
lupus erythematosus and central nervous system dysfunction. Am J Med 74:837-844, 1983
22. Attwood W, Poser CM: Neurologic complications of
Sjogren’s syndrome. Neurology 11:1034-1041, 1961
23. De la Monte SM, Hutchins GM, Gupta PK: Polymorphous meningitis with atypical mononuclear cells in
Sjogren’s syndrome. Ann Neurol 14:455-461, 1983
24. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ,
Rothfield NF, Schaller JG, Talal N, Winchester RJ: The
1982 revised criteria for the classification of systemic
lupus erythematosus. Arthritis Rheum 25: 1271-1277,
25. Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958
26. Subcommittee for Scleroderma Criteria of the American
Rheumatism Association Diagnostic and Therapeutic
Criteria Committee: Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis
Rheum 23581-590, 1980
27. Van Bijstervald OP: Diagnostic tests in the sicca syndrome. Arch Ophthalmol82: 10-14, 1969
28. Holm S: Keratocohjunctivitis sicca and the sicca syndrome. Acta Ophthalmol (suppl) 33: 1-230, 1949
29. Greenspan JS, Daniels TE, Talal N, Sylvester RA: The
histopathology of Sjogren’s syndrome in labial salivary
gland biopsies. Oral Surg 37:217-229, 1974
30. Fahey JL, McKelvey EM: Quantitative determination of
serum immunoglobulins in antibody: agarplates. J Immunol94:84-90, 1965
31. Schliep G, Felgenhauer K: Serum-CSF protein gradients, the blood-CSF barrier and the local immune response. J Neurol 218:77-96, 1978
32. Johnson KP, Arrigo SC, Nelson BJ, Ginsberg S:
Agarose electrophoresis of cerebrospinal fluid in multiple sclerosis. Neurology 27:273-277, 1977
33. Cohen SR, McKhann GM, Guarnieri M: A radioimmunoassay for myelin basic protein and its use for quantitative measurements. J Neurochem 25:371-376, 1975
34. Alexander EL, Malinow K , Lijewski JE, Jerdan MS,
Provost TT, Alexander GE: Primary Sjogren’s syndrome with central nervous system disease mimicking
multiple sclerosis. Ann Intern Med 104:323-330, 1986
35. Olsson JE, Link H: Immunoglobulin abnormalities in
multiple sclerosis. Arch Neurol 28:392-399, 1973
36. Feinglass EJ, Arnett FC, Dorsch CA, Zizic TM, Stevens
MB: Neuropsychiatric manifestations of systemic lupus
erythematosus: diagnosis, clinical spectrum, relationship to other features of the disease. Medicine (Baltimore) 55:323-329, 1976
37. Sergent JS, Lockshin MP, Klempner MS, Lipsky BA:
Central nervous system disease in systemic lupus
erythematosus: therapy and prognosis. Am J Med
58:644-654, 1975
38. Gibson T, Myers AR: Nervous system involvement in
systemic lupus erythematosus. Ann Rheum Dis 35:
398406, 1976
39. Small P, Mass MF, Kohler PF, Harbeck RJ: Central
nervous system involvement in SLE: diagnostic profile
and clinical features. Arthritis Rheum 20:869-878, 1977
40. Abel T, Gladman D, Urowitz M: Neuropsychiatric
lupus. J Rheumatol7:325-333, 1980
41. Harvey AM, Schulman L, Tumulty P, Conley CL,
Schoenrich EH: Systemic lupus erythematosus: review
of the literature and clinical analysis of 138 cases.
Medicine (Baltimore) 33:291-437, 1954
42. Johnson RT, Richardson EP: The neurological manifestations of systemic lupus erythematosus: a clinicalpathologic study of 24 cases and review of the literature.
Medicine (Baltimore) 47:337-369, 1968
43. Jaeckle KA: Cerebrospinal fluid cytomorphology in systemic lupus erythematosus with Anton’s syndrome.
Acta Cytol (Baltimore) 26532-536, 1982
44. Oehmichen M: Cerebrospinal Fluid Cytology: An Introduction and Atlas. Philadelphia, WB Saunders, 1976
45. Hauser SL, Reinherz EL, Hoban CJ, Schlossman SF,
Weiner HL: CSF cells in multiple sclerosis: monoclonal
antibody analysis and relationship to peripheral blood
T-cell subsets. Neurology 33575-579, 1983
Без категории
Размер файла
927 Кб
central, nervous, syndrome, basic, evidence, sjgren, primary, immunopathogenic, disease, system
Пожаловаться на содержимое документа