вход по аккаунту


Anti-ganglioside GM1 antibodies in guillain-barr syndrome and their relationship to Campylobacter jejuni infection.

код для вставкиСкачать
Anti-Ganghoside GM, Antibodies in
Gdain-Bar& Syndrome and Their
Relationshp to Cm$$obacter jejivni Infection
Jeremy H. Rees, MRCP,” Norman A. Gregson, PhD,? and h c h a r d A. C. Hughes, MD”
To clarify the association between Campylobucter jejuizi (Cj) infection and antibodies to ganglioside GM, (anti-GM,)
in Guillain-Barre syndrome (GBS), we have carried out a prospective case-control study of 96 patients with GRS. Cj
infection occurred in 25 (26%) patients. IgG and/or IgM anti-GM, were identified in 24 (25%)patients and in 1 of
71 (1.4%) household controls ( p < 0.001). Thirteen of the 25 (52%) Cj-positive patients had anti-GM, compared with
11 of the 71 (15%) Cj-negative patients ( p < 0.001). Neither the peak overall disability nor the 1-year disability
differed between the anti-GM,-positive and anti-GM,-negative patients. However, patients with the combination of
Cj infection and anti-GM, positivity recovered more slowly than Cj/anti-GM,-negative patients ( p = 0.05), were
more likely to have axonal degeneration, and were significantly more disabled at the end of 1 year ( p = 0.02). The
presence of Cj infection is more important than anti-GM, positivity in determining the extent of axonal involvement
and, hence, prognosis. Since the presence of anti-GM, is not a significant poor-prognostic factor, a search should be
made for other properties of Cj infection that would account for its relationship to axonal degeneration.
Rees JH, Gregson NA, Hughes RAC. Anti-ganglioside G M , antibodies in Guillain-Barri. syndrome and their
relationship to Catxpylohzi-tevjtjzmi infection. Ann Neurol 1995;38:803-816
Guillain-Barre syndrome (GBS) is a clinically defined
syndrome caused by an acute polyradiculoneuropathy
that frequently follows an infectious illness and that
is the commonest cause of acute areflexic paralysis in
developed countries. Between 20% and 302) of patients with GBS have high-titer anti-ganglioside antibodies [ 1-71, most frequently against the closely similar gangliosides G M , and GD,b.Antibodies have also
been described in GBS against ganglioside GD,, [S)
and N-acetylgalactosaminyl GD,, [9}, the latter in 6
patients with a preceding diarrheal illness. There is
controversy as to whether these antibodies are markers
of a poor prognosis due to axonal degeneration and,
in particular, whether they are associated with Cumpylobacter jejgni (Cj) infection, itself thought to be a poorprognostic factor [ lo}. Gregson and colleagues [6] reported a strong relationship between the presence of
antibody to ganglioside G M , (anti-GM,) and prolonged disability and this has been corroborated by
other authors [ 5 , 11, 12). However, neither Enders
and co-workers [ 7) nor Vriesendorp and associates
113) found a correlation between anti-GM, and Cj infection and the severity, type (axonal vs demyelinating), o r outcome of GBS.
To settle this controversy, we have prospectively examined the clinical and electrophysiological features of
a cohort of GBS patients. As reported elsewhere [ 141,
all patients have been systematically investigated for
evidence of recent Cj infection by both bacteriological
and serological methods and their clinical progress
carefully monitored. W e have measured the frequency
of anti-GM, in these patients and studied their association with C j infection, electrophysiological abnormalities, and clinical outcome.
From the Departments of *Neurology and tAnatomy and Cell Biology, UMDS, Guy’s Hospital, London, England.
Address correspondence to Dr Rees, Department of Neurology,
Royal Free Hospital, Hampstead, London NW3 ?QG, England.
Patients and Methods
Patients who fulfilled the criteria for GBS of Asbury and
Cornblath [l5] were recruited between November 1992 and
April 1994 after being reported by consultant neurologists
and physicians in district general and teaching hospitals
throughout England and Wales. Two controls were recruited
wherever possible for each patient, one from the patient’s
household (community control) and one age- and sexmatched patient admitted to the same ward as the GBS patient (hospital control). Clinical details of antecedent events
were collected and assessments of neurological disability
made according to standard rating scales used in CBS as
described by Rees and colleagues [ 141. Patients and controls
Received May 17, 1995, and in revised form Jul 13. Accepted for
publication Jul 18, 1995.
Copyright 0 1995 by the American Neurological Association
were asked to provide blooci and stool samples. T h e subsequent progress o t all patients was monitored by telephone
and those who were unable t o walk inJependently at the end
of 1 year were reexamined and testecl electrophysiologically
whenever possible. T h e study was approved by the local ethics committee and patients were seen only after informed
consent was obtained and at the tliscrction of the responsible
In addition, 21 patients with culture-positive C i enteritis
were recruited for serological studies of anti-ganglioside antibodies via their general practitioners. All patients had an uncomplicated diarrheal illness and had not developed any ncurological sequelae.
Elertrop/ Data
Electrophysiological data were gathered either from the case
notes or from nerve conduction stuliies performed by one
of us (J.H.R.). Patienrs were classified electrophysiologicaIly
as having either acutc inflammatory dcni yelinating polyradiculoneuropathy (AIDP) with or without axonal degeneration
( A D ) or acute moror axonal neuropathy ( A M A N ) or acute
motor and sensory axonal neuropathy ( A M S A N ) according
to criteria describecl by Rees a i d co-workers [ 141. Patienrs
who were not studied electrophysio1o)gically or who had insufficient evidence of demyelination or axonal degeneration
were labeled as unclassifiable.
which has been described [I71 and validated in an international comparative study [IS], was used. Wells of 96-well
ELISA plates with a hydrophobic surface (Immulon M,
Dynatech) were coated with gangliosides GM,, GD,,, or
GD,, (Sigma) by the evaporation of 50 pl of an ethanolic
solution of ganglioside 1 pg/ml and cholesterol 5 kg/ml. T h e
wells were then blocked with 200 pl of 19 bovine serum
albumin (Sigma) in high ionic strengrh phosphate-buffered
saline (HIPBS) for 2 hours at room temperature, washed
three times in HIPBS, and incubated overnight at 4°C with
5 0 pl of test serum diluted 1 : 100. After washing three times
with HIPBS, the plates were incubated at 4°C for 2 hours
with 50 p1 of extravidin-alkaline phosphatase-con jugated
rabbit anti-human IgG, IgA, or IgM (Dako) at 1 : 500 dilution. After washing in HIPBS, plates were incubated for 1
hour (1gG) o r 2 hours (IgA and IgM) with 1 mg/nil pnitrophenyl phosphate (Sigma)at 37OC and the optical densities read on an Anthos 2000 plate reader at 405 nm. Sera
with an optical density greater than three standard deviations
of the mean of the hospital controls were overlaid on thinlayer chromatograms (TLCs). as described previously [GI. A
serum was only considered to contain anti-GM, if the TLC
was also positive. T h e fine specificities of the antibodies were
determined from the pattern of lipid binding on the TLC
immunooverlay s.
Statistical Methods
Stool Sperinieiis
Stool samples were cultured dirtictly and after 24- to 48-hour
enrichment in Preston medium onto selective Curtip?/obuc-trr
media (Oxoid). Specimens were also filtered and plated onto
blood agar plates to avoid the possibility of missing Cj strains
sensitive to the antibiotics contained within the selective media (seeRees and colleagues { 141 for dctails of bacteriological
Serological Stidiej
BODIES. Sera were screened for IgG, IgA, and IgM antibodies against Cj by enzyme-linked immunosorbent assay
(ELISA) as described by Rees and colleagues [ 1 4 ] . Optical
density values of cwo standard h i a t i o n s greater than the
mean of the hospital control group ( n = 7 4 ) were taken as
the cutoff values for each serum dilution above which a sample would be called positive. Patients were diagnosed as being C j positive if Cj was cultured from stools or if anti-Cj
antibody levels were elevated in at least two classes or one
class (IgG or IgA) in the presence of a definite history of a
recent diarrheal illness (see Rees and colleagues [ 141 for further details).
96 patients, 7 1 community controls, and 65 hospital controls
(randomly chosen from the control population) were
screened for IgG, IgA, and IgM anti-ganglioside antibodies.
In addition, sera trom 22 patients with culture-positive Cj
enteritis not followed by neurological complications were
tesred for antiboclies against gangliosides G M , , GD,,, and
GD,, to determine the rate of anti-ganglioside positivity
among these patients [ 16). A modifieJ version of the ELISA,
Annals of Neurology
Vol 38
No 5
Statistical calculations were performed using Statgraphics
sofmare with two-tailed tests of significance. Differences in
proportions were tested by a x' test with Yates's correction
o r Fisher's exact test. All other comparisons between groups
were made using either Student's unpaired t test if the data
variables were normally distributed or the Mann-Whitney
U test if they were not. Univariate and multivariate logistic
analyses of poor-prognostic factors were performed with SAS
Patients atiid Controls
In the ELlSA detection of anti-GM,, the mean (SD)
optical density for the hospital control group at a serum
dilution of 1 : 100 was 0.087 (0.015) for I&, 0.079
(0.03) for IgM, and 0.07') (0.009) for IgA. Twentyfour of 96 (25y)patients (not including the ? patients
with Miller Fisher syndrome) and 1 of 71 (l.4T/i8)
community controls had IgG or IgM anti-GM, ( p =
0.00006; odds ratio, 23; 959; confidence intervals, 3476). Four patients had IgM antibodies alone, 17 IgG
antibodies alone, 1 had both, and 2 had all three
As previously reported [ 1 4 ] , 27 of 103 (267f,)patients (including the 7 patients with Miller Fisher syndrome) were diagnosed as Cj positive compared with
2 (2%,) community controls ( p < 0.0001; odds ratio,
16.3; 959$ confidence intervals, 3.6-102.8). Two of
the Cj-positive patients had Miller Fisher syndrome
and 25 had GBS. Thirteen of these 25 ( 5 2 p / o patients
had anti-ganglioside G M , antibodies compared with 11
of 71 Cj-negative GBS patients ( p = 0.0008; odds
November 1995
ratio, 5.9; 95% confidence intervals, 1.9-18.6). Nine
(699) patients had IgG antibodies alone, 2 (15%))had
IgM antibodies alone, and 2 had both. Table 1 shows
the fine specificities of the sera together with the Cjpositivity status.
Of the 22 patients with uncomplicated Cj enteritis,
2 had antibodies against gangliosides G M and G D Lb,
and one against ganglioside GD,, alone. Therefore,
less than 15% of Cj-positive patients with uncomplicated enteritis had anti-GM,. In all three cases the antibodies were IgM 1161.
Correlation with Neurological Deficit at Peak
of Illness
The anti-GM ,-positive patients had significantly less
sensory disturbance and less facial weakness than the
anti-GM,-negative patients. Other measures of severity (overall disability, arm weakness, and bulbar weakness) did not differ between the two groups (Table 2).
Correlation with Milestones of Recoileery
There were no significant differences in milestones of
recovery between the two groups as measured by the
Table 1 . Gaxglioside RractifitieA-,Immunoglobulin Classes, Campyiobacter jejuni Positii'ity Statits, arid D i ~ g n o ~ i ~
of Aiiti-GM ,-Positizie Patients with Guillain-Bawi Syndrome
Ganglioside( s)
Antibody Class
IgG(G), IgM(M), IgA(A)
G , M, A
G, M
G , M, A
Cj + v e
+ ve = Campyhbarter jejuvi positive; N = no; AlDP = acute inflammatory demyelinating polyradiculon~uropathy;Y = yes; AD = axonal
degeneration; AMAN = acute motor axonal neuroparhy; AMSAN = acute motor and sensory wonal neuropathy.
Table 2. Neurological Deficitat Peak of lllneu
Worst disability grade
Worst arm grade
Worst sensory grade
Worst facial grade
Worst bulbar grade
Anti-GM, +ve
Median Grade
Anti-GM, -ve
Median Grade
4.0 (2.0-5.0)
3.0 (1.0-4.0)
2.0 (0.0-3.0)
0.0 (0.0-3.0)
0.0 (0.0-3.0)
+ v e = positive; -ve = negative
Rees et al: Anti-GM, Antibodies in GBS
number of days to making a subjective improvement
(median [range) for anti-GM, positive and anti-GM,
negative, 12.5 days 17-33 days) and 16 days 13-56
days), respectively), to walking 5 meters across an open
space unaided (without the aid of sticks or crutches)
(median [range] for anti-GM, positive and anti-GM,
negative, 44 days [16-366 days) and 5 5 days C12-386
days), respectively), and to leaving inpatient hospital
care including rehabilitation centers (median [range]
for anti-GM, positive and anti-GM, negative, 35 days
[8-245 days) and 42 days [G-400 days], respectively).
Cowelation u i t h Disability at 1 Year
Twenty-three of the 24 (9674) anti-GM,-positive patients and 71 of the 72 (99c.i8)anti-GM,-negative patients have been followed up for l year. One patient
from each group was lost to follow-up. Two anti-GM,positive patients and 4 anti-GMI-negative patients
have died within the last year. The causes of death
were pulmonary embolism (day 72), asystolic cardiac
arrest (day ?), multiple resistant Stuphylococcus uureus
septicemia (day 89), multiorgan failure due to preexisting Eisenmenger-VSD (day 12), heart failure (day
340). and uncertain in 1 patient who had been transferred to a long-stay ward (day 240). There was a nonsignificant trend toward greater disability at the end of
1 year in the anti-GM,-positive group ( x 2 3.0; p =
0.08). If the 6 patients who died were not included in
the analysis, the difference became significant (x' 3.7;
p = 0.05). However, multivariate logistic regression of
prognostic factors predicting a poor outcome showed a
nonsignificant effect of anti-GM, positivity but a significant effect of Cj positivity, older age, becoming bed
bound within 2 days, and requirement for ventilation
(Table 3).
Electrophy.riologica/ Classifcation
The electrophysiological diagnoses were compared between the anti-GM ,-positive and anti-GM,-negative
groups. Excluding 9 cases that were unclassifiable there
was a significantly higher proportion of cases of AIDP
among the anti-GM,-negative group (n = 46 [73%,])
compared with the anti-GM,-positive group (n = 1 1
[46?6,/r7)( p = 0.03; odds ratio, 3.2; 95% confidence
intervals, 1.1-9.6). There was no significant difference
in the proportion of patients with AMAN o r AMSAN
or AIDP/AD, although a somewhat higher proportion
( 4 5 % ) of anti-GM,-positive patients with AIDP had
AD than anti-GM,-negative patients ( 2 7 9 ) ) .
Comparison of Clinical Features Between
CjlAnti-GM ,-Posititie Patients and
Cj/Anti-GM,-Nenatizie Patients
Patients were further subdivided into two groups to
see whether the combination of Cj infection and antiG M , (aGM,) positivity was associated with particular
812 Annals of Neurology
Table 3. Multirlariate Analysis of Factors Contributing
to Poor Outcome
Cj positivity
Less than 2 days
to bed bound
Odds Ratio
2.4- 106.7
aFor each incremental interval of 10 years.
Campylohacter jejiini.
clinical features. Cj/aGM,-positive patients (n = 13)
had significantly less sensory disturbance (median
grade [range] 2.0 [0.0-3.0} for both groups) ( p =
0.01) and less facial weakness (median grade [range]
0.0 10.0-2.01 for Cj/aGM ,-positive group compared
with 2.0 [0.0-3.0) for Cj/aGM,-negative group) ( p
= 0.004) at the peak of their illness than the Cj!
aGM,-negative (n = 60) group. Although there was
no difference in the overall disability grade at the nadir
of the illness, Cj/aGM,-positive patients took significantly longer to walk unaided (median [range) 85 [l63661 days) than Cj/aGM,-negative patients (median
[range) 48 [8-350) days) ( p = 0.05).
Comparison of Cj/Anti-GM,-PositiiJe Group
with CjlAnti-GM,-Negative. Cj-Positii)e/
Anti-GM,-Negative and Cj-Negatiijel
Anti-GM,-Positizie Groups
Electrophysiological diagnoses and 1-year disability
grades were compared across four groups defined by
the combinations of Cj and anti-GM, positivity (Tables
4-7) to determine whether Cj infection or anti-GM,
positivity individually was associated with a particular
electrophysiological subtype and/or prognosis. There
was a significant difference in the frequency distribution
of electrophysiological subtypes between the Cj/
aGM ,-positive group and both the C j/aGM,-negative
and the C j-negative/aGM,-positive groups but not the
Cj-positive/aGM,-negative group (Table 4). This implies that Cj infection but not aGM, positivity had an
effect on electrophysiological diagnosis. Further analysis
(Tables 5 and 6) revealed a significantly higher proportion of patients with AIDP in both Cj-negative groups.
Conversely, there was a higher proportion of patients
with AMAN/AMSAN in the Cj/aGM,-positive group
compared with the Cj/aGM,-negativegroup ( p < 0.01;
odds ratio, 22.2; 95$h confidence intervals, 1.9-591)
but not compared with the Cj-negative/aGM,-positive
group. In addition, the proportion of C j/aGM ,-positive
Vol 38 No 5 November 1995
Table 4. Electrophysiologiral Diagnoses of Groilps Aiz-ordirig to Campylobacter jejuni niid/ior Anti-GM Positiz,ity
CjiaGM, + v e
CjiaGM, -ve
Cj +ve/aGMi -ve
Cj -ve/aGM, + v e
n (7)
3 (23)
6 (46)
4 (31)
0 (0)
39 ( 7 6 )
11 (20)
1 (2)
7 (58)
3 (25)
2 (17)
0 (0)
9 (16)
Column 1 vs column 2
Column 1 vs column 3
Column 1 vs column 4
Cj = Campylubarterjejuni; aGM, = anti-GM,; +ve = positive; -ve = negative; AIDP = acute inflammatory demyelinating polyradiculoneuropathy; A D = axonal degeneration; AMAN = acute motor axonal neuropathy; AMSAN = acute motor and sensory axonal neuropathy.
Table 5 . Electrophysiofogrcal Diagnoses of C j h n t i - G M ,-Posifii]e Grorrps and CjlAiiti-GM ,-Negatiw Groiip
CjiaGM, + v e
3 (23)
6 (46)
4 (31)
Cj/aGM, -ve
n ((2)
Odds Ratio
(95% CI)
39 (76)
10.8 ( 2 . 2 , 60.4)
3.1 (0.7, 13.5)
22.2 ( 1.9, 59 1 )
0.0 (0.0, 2.7)
11 (20)
9 (16)
0 (0)
Cj = Campylubarterjejuni: aGM, = anti-GMJ; +ve = positive; -ve = negative; CI = confidence interval; AIDP = acute inflammatory
demyelinating polyradiculoneuropathy; A D = axonal degeneration; AM AN = acute motor axonal neuropathy; AMSAN = acute motor and
sensory axonal neuropathy.
Table 6. Efel.tropl?ysiologtrara( Diagnoses of Cj/Aiiti-GM ,-Pojitirv G o z i p s atid Cj-Negatir,elAriti-GM I -Positiw Group
Cj/aGM, + v e
n (9Z)
Cj -ve/aGM, + v e
n (9;)
Odds Ratio
(9556 CI)
3 (23)
6 (46)
4 (31)
0 (0)
8 (73)
3 (27)
0 (0)
0 (0)
0.1 (0.0, 0.9)
2.3 (0.3, 18.2)
Cj = Campylubarter jejknz; aGMJ = anti-GM,; +ve = positive; -ve = negative; CI = confidence interval; AIDP = acute inflammatory
demyelinating polyradiculoneuropathy; A D = axonal degeneration; AMAN = acute motor axonal neuropathy; AMSAN = acute motor and
sensory axonal neuropathy.
patients with AIDP who had A D (67%) was significantly
higher than in the Cj/aGM,-negative group (225%)( p
= 0.01; odds ratio, 7.1; 9595 confidence intervals,
1.3-44.1) (Table 5 ) but not in the Cj-negative/aGM,positive group ( p = 0.17).
The number of days to walk unaided also was significantly greater ( p = 0.03) in the Cj/aGM,-positive
group (median [range} 85 [16-3661 days) compared
with the Cj-negative/aGM,-positive group (median
[range) 29 [16-1521 days) but not compared with the
Cj-positive/aGM,-negative group.
There was significantly greater disability at 1 year
among the Cj/aGM,-positive patients (x' for trend,
5.8; p = 0.02) compared with Cj/aGM,-negative patients, but there was no difference when compared
with the other two groups (Table 7).
This is the first study to examine prospectively the
specific relationship between Cj infection, antiganglioside antibodies, and electrophysiological diagnoses in a large cohort of patients with GBS. We have
Recs et al: Anti-GM, Antibodies in GBS 813
Table. 7. One-Year Outcome
Not or slightly disabled
Moderately disabled
Severely disabled"
Cj/aGM, f v c
n ((7)
Cj/aGM, - v e
n (76)
2 (17)
2 (17)
40 (68)
14 (22)
2 (3)
4 (7)
cj +ve/aGM,
c] -ve/aGM,
5 (42)
4 (33)
3 (25)
6 (55)
4 (36)
0 (0)
1 (9)
0 (0)
for Trend
Column 1 vs column 2
Column 1 vs column 3
Column 1 vs column 4
+ ve
positive; - v e
shown that there is a significant association between
recent Cj infection and the presence of anti-GM, and
anti-GDIh.The frequency of these antibodies was 25%
in the cohort as a whole and 5 2 g ' in the 25 Cj-positive
patients with GBS. These figures confirm the previous
findings of Walsh and co-workers [2] who reported a
frequency of anti-GM, of 15'7; overall in a similar
group of 95 patients with GBS and of 5 0 g in the
14 patients with serological evidence of Cj infection.
Conversely, Enders and associates [7] found no correlation between anti-GM, and antibodies to Cj, nor between Cj, anti-GM,, prognosis, and disease type (axonal vs demyelinating). Ho and colleagues [I91 also
could not show a significant correlation between pattern of disease, anti-glycolipid antibodies, and Cj antibodies. This group, however, was studying GBS in
northern China, where an epidemic AMAN occurs in
the summer months, and they were, therefore, looking
at a different population of whom 6576, had AMAN
and were mainly children.
We have also screened a control group of 22 patients
with uncomplicated Cj enteritis for anti-GM, and
shown that they occur at a lower frequency than in
GBS patients (13% compared with 52%; odds ratio,
2.1; 95% confidence intervals, 0.5-9.9) [IG}. In addition, they were IgM in class in all 3 enteritis controls
rather than predominantly IgG as seen in the GBS
patients. This suggests that the systemic immune response to Cj in GBS differs qualitatively and quantitatively from that in simple Cj enteritis. This difference
may be due to factors in the infecting organism or in
the host response or both. Evidence for the former
hypothesis has been reported by Kuroki and associates
[20] who have shown that GBS patients are commonly
infected with the unusual Penner 19 serotype of Cj.
814 Annals of Neurology Vol 38
"Patients who were either unable to walk indepcndently or were bed/chair hound
No 5
n ("i)
ventilated at the end o f I year.
However, this has not been confirmed anywhere else
in the world and our study has shown no specific Penner serotype to be associated with GBS. We have now
collected five isolates from GBS and Miller Fisher syndrome patients of whom 2 were Penner 1, a common
serotype in the UK, 1 was Penner 64, and 2 were
nontypeable (Rees and colleagues { 141).
The fine specificities of the anti-ganglioside antibodies varied between patients. In 50%' of the anti-GM,positive group, there were antibodies that also bound
to ganglioside GD,, indicating either that the Gal(P13)GalNAc disaccharide is the epitope with which the
sera react or that there are two distinct antibody populations, one reactive against ganglioside G M , and one
against GD,,. It may be that these antibodies react
primarily with peripheral nerve glycoproteins [2 I], as
the Gal(P1-3)GalNAc structure is not specific to gangliosides. Indeed, the presence of the GM, gangliose
hapten on Cj has been demonstrated by Yuki and colleagues 122) who have shown that the oligosaccharide
of the tetrasaccharide of ganglioside G M , occurs in the lipopolysaccharide (LPS) core from a Penner 19 strain of Cj
isolated from a GBS patient. Cross-reactivity between
Cj LPS and monoclonal IgM antibodies to gangliosides
G M , and asialo-GM, from patients with chronic motor
neuropathies has also been demonstrated [23].
Anti-GM, positivity did not augur a poor prognosis;
there was no difference in the overall disability at the
peak of the illness between those patients with and
withouc anti-GM, antibodies and, although there was
a trend toward greater disability at the end of 1 year
among the anti-GM,-positive patients, this was not
statistically significant. Finally, multivariate analysis
showed that anti-GM, positivity alone was not a poor-
November 1995
prognostic factor although Cj positivity was. This is in
agreement with the findings of two recently published
studies, one from Germany 171 and one from the
United States 1131, but in contrast to previous studies
from our institution [ 2 , 61. The apparent discrepancy
may be explained by the different methodology used
in the study of Walsh and co-workers 121 who did not
confirm ELISA results with TLC-immunooverlay and
therefore may have overestimated the frequency of
anti-GM, antibodies. The study of Gregson and associates [6] only followed up patients for 12 weeks, at
which time many patients may still be unable to walk,
whereas this study monitored patients’ progress for up
to 1 year.
Comparisons between groups defined by either C j
positivity, anti-GM, positivity, neither, or both suggest
that patients with the combination of Cj and anti-GM,
positivity are more likely to have AD in addition to
AIDP or a primary axonopathy than those patients
who are negative for both Cj and anti-GM,. Further
subclassification reveals that it is the presence of Cj
infection rather than anti-GM, antibodies that is associated with the trend toward AD and axonopathy, although this could not be shown to cause a significant
difference in outcome.
These results might suggest therefore that the controversy in previous publications relating to whether
anti-GM, per se are associated with a poor outcome
may be due to the varying proportions of Cj-positive
patients within the different studies and the methods
used to detect anti-GM,. The prognosis in a group of
anti-GM,-positive patients with a high proportion of
Cj-positive patients is likely to be worse than a similar
group of patients without evidence of recent Cj infection. Anti-GM,-positive patients had less sensory disturbance and less facial weakness than anti-GM,negative patients. The reasons for this are unknown
but may depend on the distribution of antigens within
the nervous system reactive with anti-GM antibodies
of different fine specificities and their accessibility to
autoantibodies. The predilection for motor fibers may
also be due to differences in the concentration or localization of target epitopes between motor and sensory
The mechanisms whereby anti-ganglioside antibodies are generated are unknown. IgM antibodies against
GM,, asialo-GM,, GD,b, and GM, can be detected in
many normal individuals [ 2 4 ] . High titers of IgM and
IgG anti-GM, are found in patients with multifocal
motor neuropathy with conduction block 1251 and axonal motor neuropathies [26]. The production of hightiter IgG anti-GM, antibodies in GBS may be triggered
by infection with an organism such as Cj either via
cross-reactive antigens or via altered tolerance to “self”
ganglioside G M antigens induced by the formation of
a complex between the ganglioside G M , “receptor”
and the C j heat-labile enterotoxin. Cj enterotoxin may
trigger the production of anti-GM, indirectly by acting
as a superantigen that would stimulate a subset of T
cells. Certain bacteria, e.g., Staphylococitds aurem and
Stveptococcu pyogenes. produce protein toxins that are
potent activators of human T lymphocytes. Superantigens can stimulate many clones of T cells simultaneously to proliferate and some have been shown to
activate lymphocytes that react with myelin autoantigens 127). The Cj enterotoxin or cytotoxin might be a
superantigen that activates clones of T cells including
some that react with myelin. These, in turn, may then
orchestrate class switching from IgM to IgG antibody
production by B cells.
The pathogenetic relevance of anti-ganglioside antibodies has not yet been established. Yuki [ 2 8 ] has
postulated that anti-neural antibodies bind to the motor nerve terminals and cause conduction block and
subsequent axonal dysfunction. Support for this concept comes from the findings that plasma from patients
with raised anti-GM, antibodies may produce conduction block in an in vitro neuromuscular junction preparation when injected into mice 1291. However, the
exact mechanisms by which this would occur are unknown.
J.H.R. was a Medical Research Council Training Fellow. Supported
by grants from Action Research and the Guillain-Barre Syndrome
Support Group of Great Britain.
We are indebted to all the neurologists who referred their patients
to us and to the nursing staff who collected the stool specimens; to
Sister S. E. Soudain who interviewed many of the patients and to
Mr 1. Gray, Ms G. Rajagopalan, and Ms N. King for technical assistance; to D r P. Payan and Dr E. Chroni for their valuable suggestions
and expertise regarding the electrophysiological studies; and to D r
P. Griffiths who gave advice and assistance on the culturing of
Canip&bocter jejiini.
1. Koblar SA, Gregson NA, Hughes RAC, er al. Cuinpyloburter
neuropathy. Neurology 1991;4 1:1327-1328
2. Walsh FS, Cronin M, Koblar S, et al. Association between glycoconjugate antibodies and Cumpylobucter infection in patients with
Guillain-Barre syndrome. J Neuroimmunol 1991;34:43-5 1
3. Ilyas AA, Mithen FA, Dalakas MC, et al. Antibodies to acidic
glycolipids in Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy. J Neurol Sci 1992; 107:
11 1-121
4. Nobile-Orazio E, Carpo M, Meucci N , et al. Guillain-Barre
syndrome associated with high titers of anti-GM1 antibodies. J
Neurol Sci 1992;109:200-206
5. Van den Berg LH, Marrink J, de Jager AEJ, er al. Anti-GM,
antibodies in patients with Guillain-Barre syndrome. J Neurol
Neurosurg Psychiatry 1992;55:8-11
6. Gregson NA, Koblar S, Hughes RAC. Antibodies to gangliosides in Guillain-Barri. syndrome: specificity and relationship to
clinical features. Q J Med 1993;8G:111-117
7. Enders U, Karch H, Toyka KV, eta]. The spectrum of immune
responses to Campylobacter jejuni and glycoconjugates in Guillain-Barre syndrome and in other neuroimmunological disorders. Ann Neurol 1993;34:136-144
Rees et al: Anti-GM, Antibodies in GBS 815
8. Yuki N. Yoshino H, Sato S, et d. Severe acute axonal form of
Guillain-BarrC- syndrome associated with IgG anti-GD,,z antibodies. Muscle Nerve 1992;15:899-903
9. Kusunoki S. Chiba A, Kon K, et al. N-Acetylgalactosaniinyl
G D l a is a rarget molecule for serum antibody in Guillain-Barre
syndrome. Ann Neurol 1994;35:570-576
10. Kaldor J, Speed BR. Guillain-Barre syndrome and Cunipylubucrer
Jejrmi; a serological study. Br Mcd J 1984;288:1867-1870
11. Yuki N , Yoshino H, Sato S, Miyatake T. Acute axonal polyneuropathy associated with anti-GM I antibodies following Cun/p.y/nbucrrter enteritis. Neurology 1990.40,1900-1902
12. Oomes PG, Van der Ivfeche FGA. Jacobs BC. et al. Antibodies
to the ganglioside G M 1 in sera of Guillain-Barre patients recognize epitopes on Cumpylnbuctrr bacteria. Presented to the Peripheral Nerve Study Group Meeting, Arden House, New
York, 1091
13. Vriesendorp FJ, Mishu B, Blaser MJ, Koski CL. Serum antibodies to GM I , G D Ib, peripheral nerve myelin, and Cuuip~~lobui-ter
,p/uui i n patients with Guillain-Barre syndrome and controls:
correlation and prognosis. Ann Neurol 1993,34:1 30- 135
14. Rees JH, Soudain SE, Gregson NA, Hughes RAC. A prospective case control study to investigate the relationship beween
Cunzpylobuc-revjejuni infection and Guillain-Barri. syndrome. N
Engl J Med 1995 (in press)
15. Asbury AK, Cornblath DR. Assessment of current diagnostic
criteria tor Guillain-Barre syndrome. Ann Neurol 1990;
16. Recs J H , Gregson NA, Hughes RAC. Anti-ganglioside antibodies in patients with Guillain-Barri. syndrome and Cuwpi,/obuctrr
jcjurzi infection. J Infect Dis 1905;172:605
17. -Gregson N A , Jones D, Thomas PK, Willison HJ. Acute motor
neuroparhy with antibodies to GM1 ganglioside. J Neurol 199 I ;
238:44?-45 I
18. Zielasek J. Rirter G, Magi S. et al. A comparative trial of antiglycoconjugate antibody assays: IgM antibodies to G M I . J Neurol I W ; 2 4 1:475-480
816 Annals of Neurology
Vol 38
No 5
19. Ho TW,Mishu B, Li CY, et al. Guillain-Barrt syndrome in
jejuvi infection
northern China: relationship to Cun1p~/o6ui-/er
and anti-glycolipid antibodies. Brain 1995 (in press)
20. Kuroki 5, Saida T, Nukina M, er al. C'm~ip~/uburturjeJ/~,li
from patients with Guillain-BarrC syndrome belong mostly to
Penner serogroup 19 and conrain P-N-acetylglucosamirie residues. Ann Neurol 1993;33:24 3-24'
21. Thoma FP, Lee AM, Romas SN, Latov N. Monoclonal IgMs
with anti-<;al(p I-3)GalNAc activity in lower motor neurone
disease: identification of glycoprotein antigens in neural tissue
and cross-reactivity with serum immunoglobulins J Neuroiminunol 1989;23: 167-174
22. Yuki N , Taki T, lnagaki F, et al. A bacterium lipopolysaccharide
that elicits Guillain-Barre syndrome has a G M 1 ganglioside-like
structure. J Exp Med 1993;178:1771-1775
23. Wirguin I, Suturkova-Milosevic LJ, Della-Larra P. et al. Monoclonal IgM antibodies to GM1 and asialo-GM1 in chronic neuroparhies cross-react with Cnmpyhbai-ter ; ~ J N Nlipopolysaccha~
rides Ann Neurol 1994;35.698-701
2 3 . Mizutamari RK, Wiegandt H, Nores GA. Characterisation of
anti-ganglioside antibodies presrnt in normal human plasma. J
Neuroimmunol 1994;50:215-220
25. Pestronk A, Cornblath DR, llyas AA, et al. A treatable multifocal motor neuropathy with antibodies to GM 1 ganglioside. Ann
Neurol 1988;24:73-78
26. Kornberg AJ, Pestronk A, Bieser K. er al. The clinical correlates
of high-titer IgG anti-GM1 antibodies. Ann Neurol 1994;35:
2 -34-23 7
.'2 Burns J. Littlefield K, Gill G, Trotter JL. Bacterial toxin superantigens activate human T lymphocytes reactive with myelin
autoanrigens. Ann Neurol 1992;32:352-357
28. Yuki N. Pathogenesis of axonal Guillain-Barre syndrome: hypothesis. Muscle Nerve 1994;17:680-682
29. Vincent A, Roberts M, Willison H, Newsom-Davis J. Plasma
from patients with raised anti-GM 1 antibodies passively transfers conduction block. J Neuroirnmunol 1994;54:204 (Abstract)
N o v e m b e r 1995
Без категории
Размер файла
712 Кб
barry, antibodies, syndrome, gm1, guillain, ganglioside, anti, infectious, jejuni, relationships, campylobacter
Пожаловаться на содержимое документа