Download Autoimmune Responses to the Brain After Stroke Are Associated

Autoimmune Responses to the Brain After Stroke Are
Associated With Worse Outcome
Kyra J. Becker, MD; Angela J. Kalil, BA; Pat Tanzi, BSN, RNCCRC; Dannielle K. Zierath, BS;
Anna V. Savos, BS; J. Michael Gee, MS; Jessica Hadwin, BS; Kelly T. Carter, BS;
Dean Shibata, MD; Kevin C. Cain, PhD
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Background and Purpose—Immune responses to brain antigens occur after stroke, and experimental studies show that the
likelihood of developing a detrimental autoimmune response to these antigens is increased by systemic inflammation at
the time of stroke. The aim of this study was to determine if patients who developed infection in the poststroke period
would be similarly predisposed to develop autoimmune responses to central nervous system antigens.
Methods—We enrolled 114 patients within 72 hours of ischemic stroke. Clinical and demographic data were obtained, and
cellular immune responses to a panel of central nervous system antigens were assessed during the initial week and again
at Day 90. Outcome was assessed using the modified Rankin Scale.
Results—Patients who developed an infection, especially pneumonia, in the 15 days after stroke were more likely to
evidence a Th1(⫹) response to myelin basic protein and glial fibrillary acidic protein (P⫽0.019 and P⫽0.039,
respectively) at 90 days after stroke. Further, more robust Th1 responses to myelin basic protein at 90 days were
associated with a decreased likelihood of good outcome, even after adjusting for baseline stroke severity and patient age
(OR, 0.477; 95% CI, 0.244 to 0.935; P⫽0.031).
Conclusions—This study demonstrates that immune responses to brain antigens occur after stroke. Although these
responses are likely to be an epiphenomenon of ischemic brain injury, the response to myelin basic protein appears to
have clinical consequences. The potential role of postischemic autoimmune-mediated brain injury deserves further
investigation. (Stroke. 2011;42:2763-2769.)
Key Words: autoimmune 䡲 infection 䡲 MBP 䡲 outcome 䡲 stroke
A
fter stroke, there is a breakdown of the blood– brain
barrier that allows for an encounter of central nervous
system (CNS) antigens by the systemic immune system; this
encounter can occur in the injured brain as well as in the
periphery.1–5 Given that brain antigens are generally sequestered from the systemic immune system, the possibility for
developing an autoimmune response to these antigens thus
exists. In fact, studies have shown increased titers of circulating antibodies to neurofilaments and a portion of the
N-methyl-D-aspartic acid receptor in individuals with a history of stroke.6,7 Cellular immune responses to myelinassociated antigens and other brain antigens are also seen in
stroke survivors and are more robust than the responses seen
in patients with multiple sclerosis.8 –11
In an animal model of severe focal cerebral ischemia, we
showed that the predominant lymphocyte response to the
brain antigen myelin basic protein (MBP) is best characterized by antigen-specific secretion of transforming growth
factor-␤1, consistent with a T-cell regulatory response.12 If,
however, systemic inflammation was induced during the
acute stroke, animals had an increased predisposition to
develop an antigen-specific inflammatory response characterized by lymphocyte secretion of interferon-␥, consistent with
a Th1 response.12 Furthermore, those animals with a Th1(⫹)
response to MBP experienced worse 1-month outcomes.12,13
Based on these observations, we hypothesized that patients
who become infected in the immediate poststroke period
would be more likely to develop a Th1(⫹) response to brain
antigens and that this response would be associated with a
worse long-term outcome.
Materials and Methods
Research Subjects
Patients with ischemic stroke admitted to Harborview Medical
Center from September 2005 through May 2009 who were at least 18
years of age were enrolled within 72 hours of symptom onset.
Individuals with ongoing therapy for malignancy, known history of
HIV, hepatitis B or C, history of brain tumor, anemia (hematocrit
Received March 7, 2011; final revision received March 31, 2011; accepted April 25, 2011.
From the Departments of Neurology (K.J.B., A.J.K., P.T., D.K.Z., A.V.S., J.M.G.), Radiology (D.S.), and Biostatistics (K.C.C.), University of
Washington School of Medicine, Harborview Medical Center, Seattle, WA; and the University of Washington Medical Center (J.H., K.T.C.), Seattle, WA.
The online-only Data Supplement is available at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.619593/-/DC1.
Correspondence to Kyra J. Becker, MD, University of Washington School of Medicine, Box 359775, Harborview Medical Center, 325 9th Avenue,
Seattle, WA 98104-2499. E-mail [email protected]
© 2011 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org
DOI: 10.1161/STROKEAHA.111.619593
2763
2764
Stroke
October 2011
⬍35 on admission), and those taking immunomodulatory drugs were
excluded. Blood was drawn as soon as possible after stroke onset and
at 72 hours, 7 days, and 90 days after stroke onset. Blood was also
drawn from 40 volunteers to determine normative data for cellular
immune responses. The study was approved by the Institutional
Review Board; all patients or their surrogates as well as control
subjects provided informed consent.
Clinical Data
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Clinical and demographic data were collected on all patients. Stroke
severity was determined by the National Institutes of Health Stroke
Scale (NIHSS) score and outcome by the modified Rankin Scale.14,15
Details about therapeutic interventions for the treatment of stroke
were collected and there was active ascertainment of infection data.
Infection was defined as clinical symptoms of an infection (fever
and/or pyuria for urinary tract infection and fever and/or productive
cough and radiographic evidence of consolidation for pneumonia
[PNA]) and positive culture data (for both PNA and urinary tract
infection). Antibiotic therapy was used as appropriate to treat
infections; prophylactic antibiotics were not used. Stroke etiology
was determined using the Trial of ORG 10172 in Acute Stroke
Treatment criteria.16 Total infarct volume on initial diffusionweighted MRI was calculated by the ABC/2 method.17
Lymphocyte Responses
Mononuclear cells were isolated over a Ficoll gradient and frozen in
liquid nitrogen until use. Enzyme linked immunoSPOT assays were
done to detect the antigen specific secretion of interferon-␥ and
transforming growth factor-␤1 (R&D Systems). Cells were cultured
for 24 hours in 96-well plates (MultiScreen-IP; Millipore) at a
concentration of 1⫻106 per mL in media alone or with human MBP
(25 ␮g/mL; Sigma-Aldrich), human proteolipid protein (PLP; 5
␮g/mL; ABD Serotec), human neuron-specific enolase (NSE; 5
␮g/mL; Sigma-Aldrich), bovine S100 (5 ␮g/mL; Fitzgerald), or
human glial fibrillary acidic protein (GFAP 5 ␮g/mL; CalbiochemEMD BioSci) and incubated for 24 hours. The response to tetanus
toxin (TT 5 ␮g/mL; Sigma-Aldrich) was also assessed as a control.
Experiments were performed in triplicate; spots were counted using
a semiautomated system (MetaMorph). The degree of Th1 response
to each antigen is measured as the ratio of the relative increase in the
number of cells secreting interferon-␥ to the relative increase in the
number of cells secreting transforming growth factor-␤ to a given
antigen. The median response to each antigen during the first week
after stroke was determined for each patient. For the 90-day
responses, those that were ⬎75th percentile seen in control subjects
were considered to represent a Th1(⫹) response.
Laboratory Studies
Leukocyte counts were determined by the clinical hematology
laboratory. The concentrations of high-sensitivity C-reactive protein
and glucose were determined by the hospital laboratory using
standard methods. Additional plasma was immediately frozen at
⫺80°C; the concentrations of circulating cytokines (interleukin
[IL]-2, IL-6, tumor necrosis factor [TNF]-␣) were later measured
with a cytometric bead-based system (Fluorokine MAP; R&D
Systems). The sensitivity of the assays for IL-2, IL-6, and TNF-␣
was 2.23 pg/mL, 1.11 pg/mL, and 1.50 pg/mL, respectively. Values
below the limit of detection are referred to as not detected and
assigned the lowest limit of detection for statistical testing.
Statistics
Descriptive data are presented as median and interquartile range for
continuous variables and percentages for categorical variables; group
comparisons were performed using the Kruskal-Wallis H test, the
Mann-Whitney U, test or the ␹2 test statistic as appropriate. Logistic
regression was used to estimate the OR and 95% CI for the effect of
a Th1 immune response to each antigen (as a continuous measure) at
Day 90 on neurological outcome at that time point. Given the
relatively severe strokes seen in this study, good outcome was
defined as independent ambulation (modified Rankin Scale ⱕ3).
Logistic regression was also used to estimate the OR and 95% CI for
variables that predict the probability of having a Th1(⫹) immune
response to a given antigen at 90 days. Significance was set
at Pⱕ0.05.
Results
A total of 114 patients were enrolled in CASIS; none had a
history of autoimmune disease. The median age was 57 years
(range, 44 to 67 years) and the median NIHSS score was 11
(range, 4 to 19). Patients were divided into tertiles based on
initial stroke severity (the highest NIHSS score in the first 72
hours); group characteristics are presented in Table 1. The
degree of the Th1 response to the panel of brain antigens (as
well as to TT) in the first week after stroke onset is depicted
in Figure A; these responses are similar among all tertiles of
stroke severity. Four patients died in the hospital as a direct
result of their stroke/withdrawal of support. Among the
survivors, the Th1 response to PLP and MBP at 90 days was
higher in patients with more severe strokes (Figure B). In
comparison to the healthy volunteers, the Th1 response to TT
was lower in patients with stroke during the first week after
symptom onset (0.80 [0.49 to 1.79] versus 1.26 [0.92 to 1.62];
P⫽0.049), but this difference was no longer significant at
Day 90 (0.88 [0.48 to 1.80] versus 1.26 [0.92 to 1.62];
P⫽0.181). In patients with less severe strokes (NIHSS ⱕ5),
however, the response to TT was less than in the control
group early after stroke as well as at 90 days after stroke. The
responses of each tertile in comparison to the controls are
presented in Table 2.
The effect of the immune response to each antigen (as a
continuous measure) on outcome at Day 90 is presented in
Table 3. Patients with more robust Th1 responses to MBP are
less likely to experience a good outcome (modified Rankin
Scale ⱕ3). Furthermore, this effect is independent of initial
stroke severity and patient age. There are similar but less
robust effects of the immune response to PLP and GFAP on
outcome.
In this study, we defined a Th1(⫹) response to an antigen
as a response that was ⬎75% percentile of the control
response to that antigen. Patients with more severe strokes
(NIHSS ⱖ17) were more likely to develop Th1(⫹) responses
to PLP (P⫽0.026) and MBP (P⬍0.001) than patients with
less severe strokes. The primary hypothesis for this study was
that infection in the poststroke period would increase the
likelihood of an individual developing a Th1(⫹) response to
brain antigens. Table 4 shows the proportion of patients with
and without infection who had evidence of a Th1(⫹) response to each antigen at 90 days. The median response to
PLP at 90 days was also higher among patients who developed PNA in the first 15 days after stroke than those who
were not infected (1.56 [1.23 to 2.08] versus 0.90 [0.48 to
1.30]; P⫽0.025). Stroke severity was much worse among
patients who developed an infection compared to those who
did not (NIHSS⫽21 [12 to 26] versus 8 [3 to 16]; P⬍0.001).
Patients with PNA had more severe strokes (NIHSS⫽25 [21
to 30]) than patients with urinary tract infection (NIHSS⫽18
[9 to 24]) and patients without infection (NIHSS⫽8 [3 to 16];
P⬍0.001). The effect of infection on immunologic outcome
is lost after controlling for stroke severity.
Becker et al
Table 1.
Poststroke Autoimmune Responses
2765
Comparison of Patients With Stroke by Initial Stroke Severity*
NIHSS ⱕ5
(N⫽39)
NIHSS 6 –16
(N⫽36)
NIHSS ⱖ17
(N⫽39)
P
60 (46–70)
NS
Baseline demographics
Age, y
54 (40–66)
56 (43–68)
Female
11/39 (28%)
16/36 (44%)
12/39 (31%)
NS
HTN
19/39 (49%)
22/36 (61%)
20/39 (51%)
NS
DM
7/39 (18%)
8/36 (22%)
13/39 (33%)
NS
CHD
7/39 (18%)
10/36 (28%)
10/39 (26%)
NS
CABG
6/39 (15%)
7/36 (19%)
6/39 (15%)
NS
HLD
25/39 (64%)
26/36 (72%)
30/39 (77%)
NS
AF
5/39 (13%)
6/36 (17%)
5/39 (13%)
NS
13/39 (33%)
11/36 (31%)
19/39 (49%)
NS
Cardioembolic
9/39 (23%)
10/36 (28%)
12/39 (31%)
Lacunar
8/39 (20%)
3/36 (8%)
Smoker
Stroke etiology
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Atherosclerosis
Other known cause
Unknown
0
NS
0.009
3/39 (8%)
4/36 (11%)
10/39 (26%)
0.062
11/39 (28%)
9/36 (25%)
10/39 (26%)
NS
8/39 (20%)
10/36 (28%)
7/39 (18%)
NS
Stroke characteristics and treatment
Infarct volume, mL
IV tPA
0.8 (0.3–10.1)
4/39 (10%)
11.6 (1.8–32.8)
144.0 (51.1–292.7)
⬍0.001
12/36 (33%)
12/39 (31%)
0.037
Endovascular therapy
0/39
7/36 (19%)
8/39 (20%)
0.011
Hemicraniectomy
0/39
1/36 (3%)
8/39 (20%)
0.001
37.8 (37.4–38.6)
0.003
Complications
Highest temperature in first 72 h
37.3 (37.2–37.8)
37.6 (37.2–37.9)
Infection by day 15
4/39 (10%)
4/36 (11%)
PNA by day 15
21/38 (55%)
⬍0.001
12/38 (32%)
⬍0.001
0/39
0/36
hsCRP per mg/L†
2.4 (1.5–6.2)
10.6 (4.6–40.0)
66.6 (19.2–112.2)
⬍0.001
⬍0.001
Concentrations of inflammatory markers
IL-6 pg/mL†
1.0 (ND-1.9)
2.2 (0.8–5.8)
9.5 (4.2–25.6)
TNF-␣ pg/mL†
1.2 (ND-2.8)
1.9 (0.7–3.0)
2.1 (ND-3.7)
NS
IL-2 pg/mL†
ND (ND-ND)
ND (ND-ND)
ND (ND-2.6)
0.005
Laboratory variables
Glucose, mg/dL†
119 (107–143)
123 (107–151)
143 (129–171)
Cortisol, ␮g/dL†
9.8 (7.6–13.8)
13.9 (10.9–16.7)
20.2 (13.8–27.4)
19.0 (13.0–28.2)
20.0 (11.5–40.5)
22.5 (16.5–32.0)
NS
8.1 (6.0–10.1)
10.3 (8.7–13.8)
12.1 (10.5–15.0)
⬍0.001
ACTH, pg/mL†
WBCs, thousands/mL†
0.002
⬍0.001
PMNs, thousands/mL†
3.9 (3.0–5.2)
5.9 (4.5–7.6)
8.1 (6.3–9.7)
⬍0.001
Lymphs, thousands/mL‡
1.7 (1.3–2.1)
1.4 (1.2–1.7)
1.2 (0.9–1.5)
0.002
NIHSS indicates National Institutes of Health Stroke Scale; HTN, hypertension; DM, diabetes mellitus; CHD, coronary heart disease;
CABG, coronary artery bypass grafting; HLD, hyperlipidemia; AF, atrial fibrillation; IV tPA, intravenous tissue plasminogen activator;
PNA, pneumonia; hsCRP, high-sensitivity C-reactive protein; IL, interleukin; TNF, tumor necrosis factor; ACTH, adrenocorticotrophic
hormone; WBCs, white blood cells; PMNs, polymorphonuclear cells; lymphs, lymphocytes; ND, not detected; NS, nonsignificant.
*Data are presented as either the median and interquartile range or the proportion. Statistics are by Kruskal-Wallis H test or ␹2
as appropriate.
†The highest value in the first 72 h.
‡Lowest value in the first 72 h.
Univariate predictors of developing a Th1 response ⬎75th
percentile of the control population at Day 90 for each of the
antigens studied (a Th1[⫹] response) are shown in Supplemental Table I (http://stroke.ahajournals.org). Stroke severity
was one of the most important predictors of developing a
Th1(⫹) response to most antigens and was more predictive
than infarct volume. For each 1-point increase in the NIHSS
score, there was nearly a 10% increase in the odds of
developing a Th1(⫹) response to MBP. There were trends
toward increased risk of developing Th1(⫹) response to TT
and GFAP among patients undergoing hemicraniectomy;
these trends no longer existed after controlling for stroke
2766
Stroke
October 2011
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Figure. Median Th1 responses (Y-axis) to TT and brain antigens in the first week after stroke onset (A) do not differ based on tertile of
initial stroke severity (X-axis). By Day 90, the responses to PLP and MBP are more robust in patients with severe strokes (B). Statistics
are by Kruskal-Wallis H; *P⬍0.05, **P⬍0.001. TT indicates tetanus toxin; PLP, proteolipid protein; MBP, myelin basic protein; NSE,
neuron specific enolase; GFAP, glial fibrillary acidic protein.
severity. Curiously, higher numbers of lymphocytes were
associated with a decreased risk of developing a Th1(⫹)
response to PLP and GFAP. Not surprisingly, infection (and
fever) in the first 15 days after stroke onset was associated
with increased risk of developing a Th1(⫹) response to TT,
PLP, MBP, and GFAP. Finally, Gram-positive infections
were more likely to be associated with a subsequent Th1(⫹)
response (likely because the preponderance of Gram-positive
infections were PNAs).18
Discussion
After stroke, lymphocytes infiltrate the ischemic brain allowing for contact with CNS antigens from numerous different
cell types (neurons, astrocytes, oligodendrocytes), which are
normally sequestered from the peripheral immune system.1–3
Furthermore, there is an increase in the concentration of
antigens such as MBP, NSE, S-100, and GFAP in the
systemic circulation after stroke, which allows for lymphocyte encounter with these antigens in peripheral lymphoid
Table 2. Differences Between the Th1 Response to Each Antigen for Each Tertile of Stroke Severity and Controls After Stroke Onset
and at Day 90*
Control
Subjects
(N⫽39)
Tertile 1
NIHSS ⱕ5
(N⫽38)
P
Tertile 2
NIHSS 6 –16
(N⫽36)
P
Tertile 3
NIHSS ⱖ17
(N⫽36)
P
TT
1.26 (0.92–1.62)
PLP
0.85 (0.63–1.85)
0.87 (0.47–1.33)
0.020
1.04 (0.73–1.45)
NS
0.70 (0.45–2.09)
0.073
1.05 (0.60–2.17)
NS
0.95 (0.58–1.40)
NS
1.08 (0.56–1.67)
MBP
1.02 (0.73–1.47)
1.04 (0.71–1.73)
NS
NS
0.88 (0.63–1.50)
NS
1.04 (0.70–1.66)
NSE
0.92 (0.55–1.38)
NS
1.12 (0.70–1.92)
NS
0.86 (0.44–1.56)
NS
0.98 (0.72–1.55)
NS
GFAP
S-100B
1.11 (0.51–1.46)
1.00 (0.55–1.69)
NS
0.80 (0.45–1.31)
NS
1.12 (0.78–1.57)
NS
1.00 (0.60–1.42)
0.98 (0.71–1.60)
NS
0.76 (0.50–1.46)
NS
1.20 (0.82–1.47)
NS
TT
1.26 (0.92–1.62)
0.78 (0.35–1.88)
0.033
0.83 (0.52–1.49)
0.092
1.38 (0.52–1.97)
NS
PLP
0.85 (0.63–1.85)
0.89 (0.65–1.30)
NS
0.78 (0.43–1.55)
NS
1.40 (0.80–2.31)
0.115
MBP
1.02 (0.73–1.47)
0.81 (0.54–1.06)
0.060
0.88 (0.57–1.42)
NS
1.68 (0.71–2.60)
0.094
NSE
0.92 (0.55–1.38)
0.68 (0.57–1.19)
NS
0.93 (0.51–1.41)
NS
1.00 (0.72–1.74)
NS
GFAP
1.11 (0.51–1.46)
0.78 (0.40–1.05)
0.062
0.98 (0.48–1.44)
NS
0.96 (0.53–2.13)
NS
S-100B
1.00 (0.60–1.42)
0.86 (0.53–1.54)
NS
0.77 (0.45–1.97)
NS
1.65 (0.69–3.43)
0.079
Within first week
Day 90
NIHSS indicates National Institutes of Health Stroke Scale; TT, tetanus toxin; PLP, proteolipid protein; MBP, myelin basic protein; NSE, neuron-specific enolase;
GFAP, glial acidic fibrillary acidic protein; NS, signifies P⫾0.200.
*Data are displayed as median (interquartile range); statistics are by Mann-Whitney U test (stroke patients vs control subjects).
Becker et al
Table 3. The Effect of the Immune Response to Each Antigen
at 90 Days (as a Continuous Variable) on 90-Day Outcome*
90-Day Outcome
OR
P
At 90 d
TT
(N⫽80)
PLP
(N⫽79)
MBP
(N⫽80)
NSE
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
(N⫽80)
GFAP
(N⫽77)
Unadjusted
1.015 (0.866–1.190)
NS
Adjusted for NIHSS
1.029 (0.876–1.208)
NS
Adjusted for NIHSS/age
1.023 (0.868–1.205)
NS
Unadjusted
0.644 (0.426–0.973)
0.037
Adjusted for NIHSS
0.702 (0.460–1.071)
0.100
Adjusted for NIHSS/age
0.677 (0.438–1.045)
0.078
Unadjusted
0.437 (0.251–0.763)
0.004
Adjusted for NIHSS
0.469 (0.240–0.919)
0.027
Adjusted for NIHSS/age
0.477 (0.244–0.935)
0.031
Unadjusted
0.833 (0.505–1.373)
NS
Adjusted for NIHSS
1.000 (0.556–1.801)
NS
Adjusted for NIHSS/age
1.060 (0.590–1.904)
NS
Unadjusted
0.669 (0.442–1.012)
0.057
Adjusted for NIHSS
0.783 (0.497–1.232)
NS
Unadjusted
1.001 (0.775–1.294)
NS
Adjusted for NIHSS
1.171 (0.829–1.654)
NS
Adjusted for NIHSS/age
NS
Adjusted for NIHSS/age
S-100
(N⫽73)
mRS indicates modified Rankin Scale; TT, tetanus toxin; PLP, proteolipid
protein; MBP, myelin basic protein; NSE, neuron-specific enolase; GFAP, glial
fibrillary acidic protein; NIHSS, National Institutes of Health Stroke Scale; NS
signifies P⫾0.200; OR, odds ratio.
*The ORs are either unadjusted or adjusted for baseline NIHSS score or
baseline NIHSS score and age.
organs.5,19 Given this lymphocyte contact with novel CNS
antigens, irrespective of site, it is thus possible for an
(auto)immune response to occur to these antigens. Indeed,
antibodies and cellular immune responses to brain antigens
are documented in stroke survivors.6 –11 Our observation that
Th1(⫹) responses to CNS antigens like MBP occur after
stroke is thus not novel, but the finding that these responses
may have pathological consequences is.
In experimental studies of severe stroke, we showed that
Th1(⫹) responses to MBP were uncommon under usual
Table 4. Proportion of Patients With a Th1(ⴙ) Response (Response >75th Percentile of the
Control Population) to a Given Antigen Based on Infection History*
Any Infection by Day 15?
Yes
2767
circumstances; the tendency to develop a Th1(⫹) response to
MBP, however, could be increased by induction of a systemic
inflammatory response with lipopolysaccharide, a component
of the Gram-negative bacterial cell wall, at the time of
stroke.12,13 The systemic inflammatory response induced by
lipopolysaccharide was intended to mimic the response that
occurs during infection. Based on our animal data, we
hypothesized that patients who became infected in the immediate poststroke period would be more likely to develop
Th1(⫹) responses to brain antigens. The findings from this
study support our hypothesis that infection would increase the
likelihood of developing a Th1(⫹) response to brain antigens
given that a greater proportion of patients with poststroke
infection (especially PNA) developed Th1(⫹) responses to
the myelin-associated antigens (MBP and PLP) and the glial
antigen GFAP. The fact that infections with Gram-positive
organisms were associated with increased risk for developing
Th1(⫹) responses in this study likely relates to the fact that
patients with urinary tract infections, which were caused
primarily by Gram-negative pathogens, had less severe
strokes than patients with PNA, which were predominantly
related to Gram-positive organisms.18
Potentially the most important observation in this study is
that the more robust the immune response to MBP (and to a
lesser extent, PLP and GFAP) at 90 days, the greater the
chances of a poor outcome at this time point. A similar
association between a Th1(⫹) response to MBP and poor
outcome after stroke was demonstrated in animal studies.12,13,20 MBP is a myelin-associated protein found in
oligodendrocytes within the CNS, and PLP is the most
abundant protein in myelin. Both MBP and PLP are implicated as autoantigens in multiple sclerosis and can induce
experimental autoimmune encephalomyelitis.21 We also observed a trend toward worse outcome in patients with more
robust responses to GFAP. Both humoral responses and
cellular immune responses to GFAP have been described in
patients with dementia.22–24 Immune responses to neuronspecific enolase, S-100, and TT, on the other hand, were not
predictive of outcome.
Given that severe strokes and infections are both associated
with the development of more robust immune responses to TT,
PLP, MBP, and GFAP, it is possible to argue that the responses
mRS ⱕ3
Response to
Poststroke Autoimmune Responses
PNA by Day 15?
No
P
Yes
No
P
TT
11/22 (50%)
14/59 (24%)
0.023
5/10 (50%)
14/59 (24%)
0.085
PLP
6/22 (27%)
10/58 (17%)
NS
4/10 (40%)
10/58 (17%)
0.100
MBP
9/22 (41%)
14/59 (24%)
0.127
6/10 (60%)
14/59 (24%)
0.019
NSE
5/23 (22%)
14/58 (24%)
NS
4/11 (36%)
14/58 (24%)
NS
GFAP
8/22 (36%)
11/56 (20%)
0.122
5/10 (50%)
11/56 (20%)
0.039
S-100B
8/21 (38%)
18/53 (34%)
NS
3/9 (33%)
18/53 (34%)
NS
PNA indicates pneumonia; TT, tetanus toxoid; PLP, proteolipid protein; MBP, myelin basic protein; NSE, neuron-specific
enolase; GFAP, glial fibrillary acidic protein; NS signifies Pⱖ0.200.
*Statistics are by ␹2 and compare patients with infection (any vs no infection or PNA vs no infection) with those without
infection.
2768
Stroke
October 2011
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
to these antigens are merely a marker for patients with severe
stroke and hence worse outcome. The fact that more robust
immune responses to TT were not associated with worse
outcome, however, tends to dispute this argument as does the
fact that the predictive value of the immune response on
outcome, at least for MBP, was largely unaltered after controlling for stroke severity. Furthermore, the fact that the responses
to NSE and S-100 were not substantially related to stroke
severity, infection, or outcome suggests that the observed responses to PLP, MBP, and GFAP are more than just a nonspecific inflammation in patients with severe stroke.
Molecular mimicry and bystander activation are concepts
that have been proposed to explain how infection can induce
autoimmunity.25 For molecular mimicry, it is believed that
the immune response to the pathogen cross-reacts with a
self-antigen; for bystander activation, it is believed that the
infection/inflammation leads to the expression of the necessary costimulatory molecules (“danger signals”) that activate
lymphocytes leading to an immune response to antigens in
the vicinity of the “danger signals.” Our previous experimental work showed that administration of lipopolysaccharide at
the time of stroke led to upregulation of B7.1, an important
costimulatory molecule, in the brain.12 There are also abundant data showing that lipopolysaccharide and other inflammatory stimuli lead to upregulation of B7.1 in monocytes and
dendritic tissue in the periphery.26,27 It is thus plausible that
an infection like PNA provokes enough of an inflammatory
response to upregulate the expression of costimulatory molecules in either peripheral lymphoid organs or in the brain to
create a microenvironment that promotes bystander activation
of lymphocytes encountering CNS antigens. The finding that
increasing lymphocyte numbers were associated with a decreased likelihood of developing a Th1(⫹) response to PLP
and GFAP is counterintuitive but may be explained by the
profound effect of stroke severity on lymphocyte numbers
(Table 1) and the fact that stroke severity is such a potent
predictor of the Th1(⫹) response to these antigens.
Limitations of this study include the relatively small
sample size and the fact that we assessed responses to only a
small number of potential antigens. Furthermore, our control
population was not fully characterized and matched to the
patients with stroke; for the most part, the control subjects
were largely young and healthy with no history of stroke.
Larger studies will need to be done to confirm our results and
determine whether stroke severity or infection is more important in driving the autoimmune responses to CNS antigens
and to establish normative values for larger cohorts of healthy
volunteers. These studies could also assess the response to
portions of MBP and PLP with defined antigenicity and
evaluate the Th17 response as well as the Th1 response to
these antigens. Finally, new approaches to identifying novel
autoantigens could also be considered in future studies.
In summary, this study again confirms that cellular immune responses to brain antigens occur after stroke. More
importantly, we showed that the response to some of these
antigens, MBP in particular, predicts long-term outcome. The
risk of developing a Th1(⫹) response to MBP is increased by
PNA in the first 15 days after stroke, consistent with
observations from our animal model. The development of a
detrimental autoimmune response to the brain after stroke
may help to explain why infection is an independent predictor
of poor outcome after stroke.
Sources of Funding
This research was funded by National Institute of Neurological
Disorders and Stroke 5R01NS049197.
Disclosures
None.
References
1. Schroeter M, Jander S, Witte OW, Stoll G. Local immune responses in
the rat cerebral cortex after middle cerebral artery occlusion. J Neuroimmunol. 1994;55:195–203.
2. Jander S, Kraemer M, Schroeter M, Witte OW, Stoll G. Lymphocytic
infiltration and expression of intercellular adhesion molecule-1 in photochemically induced ischemia of the rat cortex. J Cereb Blood Flow
Metab. 1995;15:42–51.
3. Braun JS, Jander S, Schroeter M, Witte OW, Stoll G. Spatiotemporal
relationship of apoptotic cell death to lymphomonocytic infiltration in
photochemically induced focal ischemia of the rat cerebral cortex. Acta
Neuropathol (Berl). 1996;92:255–263.
4. Campanella M, Sciorati C, Tarozzo G, Beltramo M. Flow cytometric
analysis of inflammatory cells in ischemic rat brain. Stroke. 2002;33:
586 –592.
5. Jauch EC, Lindsell C, Broderick J, Fagan SC, Tilley BC, Levine SR.
Association of serial biochemical markers with acute ischemic stroke: the
National Institute of Neurological Disorders and Stroke recombinant
tissue plasminogen activator stroke study. Stroke. 2006;37:2508 –2513.
6. Bornstein NM, Aronovich B, Korczyn AD, Shavit S, Michaelson DM,
Chapman J. Antibodies to brain antigens following stroke. Neurology.
2001;56:529 –530.
7. Dambinova SA, Khounteev GA, Izykenova GA, Zavolokov IG, Ilyukhina
AY, Skoromets AA. Blood test detecting autoantibodies to N-methyl-Daspartate neuroreceptors for evaluation of patients with transient ischemic
attack and stroke. Clin Chem. 2003;49:1752–1762.
8. Youngchaiyud U, Coates AS, Whittingham S, Mackay IR. Cellularimmune response to myelin protein: absence in multiple sclerosis and
presence in cerebrovascular accidents. Aust N Z J Med. 1974;4:535–538.
9. Kallen B, Nilsson O, Thelin C. Effect of encephalitogenic protein on
migration in agarose of leukocytes from patients with multiple sclerosis.
A longitudinal study of patients with relapsing multiple sclerosis or with
cerebral infarction. Acta Neurol Scand. 1977;55:47–56.
10. Wang WZ, Olsson T, Kostulas V, Hojeberg B, Ekre HP, Link H. Myelin
antigen reactive T cells in cerebrovascular diseases. Clin Exp Immunol.
1992;88:157–162.
11. Rocklin RE, Sheremata WA, Feldman RG, Kies MW, David JR. The
Guillain-Barré syndrome and multiple sclerosis. In vitro cellular
responses to nervous-tissue antigens. N Engl J Med. 1971;284:803– 808.
12. Becker KJ, Kindrick DL, Lester MP, Shea C, Ye ZC. Sensitization to
brain antigens after stroke is augmented by lipopolysaccharide. J Cereb
Blood Flow Metab. 2005;25:1634 –1644.
13. Zierath D, Thullbery M, Hadwin J, Gee JM, Savos A, Kalil A, Becker KJ.
CNS immune responses following experimental stroke. Neurocrit Care.
2010;12:274 –284.
14. Lyden P, Lu M, Jackson C, Marler J, Kothari R, Brott T, Zivin J.
Underlying structure of the national institutes of health stroke scale:
results of a factor analysis. NINDS tPA stroke trial investigators. Stroke.
1999;30:2347–2354.
15. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients.
Stroke. 1988;19:604 – 607.
16. Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL,
et al. Classification of subtype of acute ischemic stroke. Definitions for
use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute
Stroke Treatment. Stroke. 1993;24:35– 41.
17. Sims JR, Gharai LR, Schaefer PW, Vangel M, Rosenthal ES, Lev MH, et
al. ABC/2 for rapid clinical estimate of infarct, perfusion, and mismatch
volumes. Neurology. 2009;72:2104 –2110.
18. Tanzi P, Cain K, Kalil A, Zierath D, Savos A, Gee JM, et al. Post-stroke
infection: a role for IL-1ra? Neurocrit Care. 2010 Dec 12 [Epub ahead of
print].
Becker et al
19. Herrmann M, Vos P, Wunderlich MT, de Bruijn CH, Lamers KJ. Release
of glial tissue-specific proteins after acute stroke: a comparative analysis
of serum concentrations of protein S-100b and glial fibrillary acidic
protein. Stroke. 2000;31:2670 –2677.
20. Gee JM, Kalil A, Thullbery M, Becker KJ. Induction of immunologic
tolerance to myelin basic protein prevents central nervous system autoimmunity and improves outcome after stroke. Stroke. 2008;39:
1575–1582.
21. Mannie M, Swanborg RH, Stepaniak JA. Experimental autoimmune
encephalomyelitis in the rat. Curr Protoc Immunol. 2009;85:
15.2.1–15.2.15.
22. Ishida K, Kaneko K, Kubota T, Itoh Y, Miyatake T, Matsushita M, et al.
Identification and characterization of an anti-glial fibrillary acidic protein
antibody with a unique specificity in a demented patient with an autoimmune disorder. J Neurol Sci. 1997;151:41– 48.
Poststroke Autoimmune Responses
2769
23. Mecocci P, Parnetti L, Romano G, Scarelli A, Chionne F, Cecchetti R, et
al. Serum anti-GFAP and anti-S100 autoantibodies in brain aging, Alzheimer’s disease and vascular dementia. J Neuroimmunol. 1995;57:
165–170.
24. Nakamura K, Takeda M, Tanaka T, Tanaka J, Kato Y, Nishinuma K, et
al. Glial fibrillary acidic protein stimulates proliferation and immunoglobulin synthesis of lymphocytes from Alzheimer’s disease patients.
Methods Find Exp Clin Pharmacol. 1992;14:141–149.
25. Fujinami RS, von Herrath MG, Christen U, Whitton JL. Molecular
mimicry, bystander activation, or viral persistence: infections and autoimmune disease. Clin Microbiol Rev. 2006;19:80 –94.
26. Schmittel A, Scheibenbogen C, Keilholz U. Lipopolysaccharide effectively up-regulates B7–1 (CD80) expression and costimulatory function
of human monocytes. Scand J Immunol. 1995;42:701–704.
27. Harris NL, Ronchese F. The role of B7 costimulation in T-cell immunity.
Immunol Cell Biol. 1999;77:304 –311.
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Autoimmune Responses to the Brain After Stroke Are Associated With Worse Outcome
Kyra J. Becker, Angela J. Kalil, Pat Tanzi, Dannielle K. Zierath, Anna V. Savos, J. Michael
Gee, Jessica Hadwin, Kelly T. Carter, Dean Shibata and Kevin C. Cain
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Stroke. 2011;42:2763-2769; originally published online July 28, 2011;
doi: 10.1161/STROKEAHA.111.619593
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2011 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/42/10/2763
Data Supplement (unedited) at:
http://stroke.ahajournals.org/content/suppl/2011/07/29/STROKEAHA.111.619593.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/
SUPPLEMENTAL MATERIAL
Supplemental Table. Predictors of a TH1(+) response to each antigen at 90 days after stroke. Laboratory
data represent the most extreme value for the first 72 hours after stroke onset. Results are expressed as
the odds ratio (OR) and 95% confidence interval (CI).
TT
antigen:
PLP
OR
P
1.008
NS
MBP
OR
P
1.006
NS
NSE
OR
P
1.022
NS
GFAP
OR
P
1.002
NS
S100
OR
P
1.017
NS
OR
P
1.012
NS
baseline demographics
age (per year)
gender (female)
NIHSS score
(0.972-1.046)
0.526
(0.181-1.527)
1.056
(1.003-1.112)
NS
0.039
(0.965-1.049)
0.223
(0.047-1.065)
1.070
(1.008-1.136)
0.060
0.025
(0.983-1.062)
0.605
(0.206-1.773)
1.099
(1.038-1.164)
NS
0.001
(0.964-1.043)
0.649
(0.206-2.043)
1.023
(0.968-1.080)
NS
NS
(0.977-1.059)
0.483
(0.142-1.643)
1.069
(1.011-1.131)
NS
0.020
(0.974-1.051)
1.059
(0.387-2.897)
1.052
(0.998-1.109)
NS
0.059
stroke etiology
cardioembolic
1.286
(0.460-3.594)
0.347
lacunar
(0.040-3.048)
atherosclerosis
other known cause
1.136
(0.260-4.961)
0.667
(0.227-1.955)
1.591
unknown
(0.533-4.751)
NS
NS
NS
NS
NS
0.789
(0.226-2.762)
0.644
(0.072-5.769)
1.163
(0.218-6.221)
1.533
(0.486-4.840)
0.824
(0.206-3.303)
NS
NS
NS
NS
NS
1.148
(0.398-3.310)
1.010
(0.182-5.615)
3.750
(0.907-15.497)
0.570
(0.184-1.769)
0.662
(0.193-2.274)
NS
NS
0.068
NS
NS
0.945
(0.295-3.027)
2.719
(0.551-13.412)
0.924
(0.175-4.876)
0.750
(0.237-2.372)
0.914
(0.261-3.202)
NS
NS
NS
NS
NS
1.354
(0.437-4.195)
1.271
(0.226-7.153)
0.354
(0.041-3.032)
1.051
(0.345-3.205)
0.944
(0.267-3.337)
NS
NS
NS
NS
NS
1.425
(0.510-3.987)
0.917
(0.156-5.375)
0.583
(0.109-3.121)
0.600
(0.201-1.788)
1.596
(0.516-4.941)
NS
NS
NS
NS
NS
stroke characteristics and treatment
infarct volume (per 10 cc)
1.038
(0.998-1.080)
2.062
IV tPA
(0.732-5.815)
hemicraniectomy
4.417
(0.965-20.213)
0.066
0.171
0.056
1.037
(0.994-1.081)
1.364
(0.411-4.523)
2.723
(0.577-12.858)
0.089
NS
NS
1.038
(0.998-1.080)
0.729
(0.232-2.294)
1.590
(0.347-7.277)
0.063
NS
NS
1.000
(0.957-1.046)
0.706
(0.205-2.430)
1.098
(0.203-5.950)
NS
NS
NS
1.034
(0.993-1.077)
0.960
(0.297-3.097)
3.667
(0.819-16.415)
0.101
NS
0.089
1.029
(0.990-1.071)
2.787
(0.981-7.918)
2.000
(0.456-8.763)
0.150
0.054
NS
inflammatory cytokine concentrations (highest by 72 hours)
IL-6 (per pg/mL)
hsCRP (per 10 mg/L)
TNF-α (per pg/mL)
IL-2 (per pg/mL)
1.010
(0.993-1.028)
1.070
(0.975-1.016)
1.226
(0.970-1.549)
3.396
(0.895-12.884)
NS
0.153
0.089
0.072
1.003
(0.985-1.021)
1.064
(0.963-1.177)
1.042
(0.791-1.374)
1.235
(0.706-2.158)
NS
NS
NS
NS
1.002
(0.986-1.019)
1.097
(0.998-1.207)
1.100
(0.870-1.392)
1.090
(0.627-1.893)
NS
0.056
NS
NS
1.000
(0.982-1.019)
0.973
(0.872-1.085)
1.010
(0.773-1.319)
0.914
(0.436-1.914)
NS
NS
NS
NS
1.017
(0.996-1.039)
1.063
(0.965-1.170)
1.066
(0.835-1.362)
1.335
(0.741-2.403)
0.109
NS
NS
NS
0.997
(0.979-1.015)
1.026
(0.990-1.132)
1.063
(0.810-1.395)
2.245
(0.670-7.421)
NS
NS
NS
0.185
laboratory values (highest by 72 hours)
blood glucose (per 10 mg/dL)
WBCs (per thou/mL)
PMNs (per thou/mL)
lymphs* (per thou/mL)
0.957
(0.870-1.051)
1.014
(0.906-1.135)
1.088
(0.925-1.279)
0.566
(0.215-1.489)
NS
NS
NS
NS
1.014
(0.923-1.114)
0.981
(0.856-1.124)
1.098
(0.913-1.322)
0.188
(0.043-0.813)
NS
NS
NS
0.025
1.045
(0.962-1.134)
1.070
(0.955-1.198)
1.157
(0.979-1.368)
0.729
(0.286-1.859)
NS
NS
0.087
NS
1.014
(0.928-1.109)
0.975
(0.860-1.106)
0.963
(0.805-1.152)
0.933
(0.367-2.375)
NS
NS
NS
NS
1.040
(0.954-1.133)
1.018
(0.900-1.152)
1.164
(0.969-1.399)
0.220
(0.059-0.825)
NS
NS
0.105
0.025
0.958
(0.871-1.053)
0.975
(0.869-1.094)
0.996
(0.845-1.174)
0.726
(0.295-1.789)
NS
NS
NS
NS
complications
highest temp in 1st 72 hrs
(per OC)
2.394
(1.111-5.158)
st
highest temp in 1 week
1.347
(per C)
(0.787-2.307)
O
infection by day 15
PNA by day 15
gram positive infection
gram negative infection
3.214
(1.149-8.992)
2.550
(0.666-9.769)
5.056
(1.323-19.313)
1.895
(0.580-6.193)
0.026
NS
0.026
0.172
0.018
NS
3.124
(1.306-7.469)
1.481
(0.798-2.749)
1.800
(0.565-5.739)
3.222
(0.787-13.193)
2.714
(0.685-10.759)
0.619
(0.124-3.094)
0.010
NS
NS
0.104
0.155
NS
1.678
(0.798-3.527)
1.534
(0.881-2.670)
2.225
(0.786-6.297)
4.765
(1.202-18.893)
3.741
(1.013-13.817)
1.011
(0.282-3.618)
0.172
0.130
0.132
0.026
0.048
NS
1.068
(0.529-2.154)
1.334
(0.729-2.44)
0.873
(0.274-2.782)
2.095
(0.541-8.120)
2.095
(0.541-8.120)
0.443
(0.091-2.169)
NS
NS
NS
NS
NS
NS
3.263
(1.390-7.661)
1.823
(0.986-3.371)
2.338
(0.786-6.956)
3.857
(0.978-15.208)
4.985
(1.315-18.891)
1.307
(0.358-4.774)
0.007
0.056
0.127
0.054
0.018
NS
1.294
(0.633-2.645)
0.952
(0.567-1.598)
1.197
(0.419-3.413)
0.913
(0.209-3.996)
2.048
(0.534-7.859)
0.691
(0.193-2.467)
NS
NS
NS
NS
NS
NS
TT=tetanus toxoid, PLP=proteolipid protein, MBP=myelin basic protein, NSE=neuron specific enolase, GFAP=glial fibrillary acidic protein, NIHSS=NIH stroke
scale, IV tPA=intravenous tissue plasminogen activator, IL=interleukin, TNF=tumor necrosis factor, WBCs=white blood cells, PMNs=polymorphonuclear cells,
lymphs=lymphocytes, temp=temperature, PNA=pneumonia. *denotes lowest value by 72 hours. NS signifies P≥0.020.