close

Вход

Забыли?

вход по аккаунту

?

Deactivation of vascular endothelium by monoclonal antitumor necrosis factor ╨Ю┬▒ antibody in rheumatoid arthritis.

код для вставкиСкачать
ARTHRITIS & RHEUMATISM
Vol. 39, No. 7, July 1996, pp 1082-1091
0 1996, American College of Rheumatology
1082
DEACTIVATION OF VASCULAR ENDOTHELIUM BY
MONOCLONAL ANTI-TUMOR NECROSIS FACTOR a ANTIBODY IN
RHEUMATOID ARTHRITIS
EWA M. PALEOLOG, MARY HUNT, MICHAEL J. ELLIOTT, MARC F E L D M A " ,
RAVINDER N. MAINI, and JAMES N. WOODY
Objective. To assess whether monoclonal antibody to tumor necrosis factor a (TNFa) reduces endothelial activation in rheumatoid arthritis (RA).
Methods. Levels of serum E-selectin, intercellular
adhesion molecule 1 (ICAM-l), and vascular cell adhesion molecule 1 (VCAM-l), and circulating leukocytes
(differential counts) were measured in RA patients
before and up to 4 weeks after infusion of either placebo
or chimeric anti-TNFa antibody cA2 (1 or 10 mg/kg).
Results. Treatment with anti-TNFa decreased
serum E-selectin and ICAM-1 levels, with the earliest
detectable changes observed on days 1-3 after
anti-TNFa infusion. No effect on VCAM-1 levels was
detected. In parallel, there was a rapid and sustained
increase in circulating lymphocytes. The extent of the
decrease in serum E-selectin and ICAM-1 levels and the
increase in lymphocyte counts was significantly higher
(P I 0.05) in patients in whom a clinical benefit of
anti-TNFa was observed (120% response, by Paulus
criteria, a t week 4) compared with that in patients who
failed to respond to anti-TNFa at this time point.
Conclusion. We propose that decreased serum
levels of adhesion molecules may reflect diminished
activation of endothelial cells in the synovial microvasSupported by a grant from the Arthritis and Rheumatism
Council of Great Britain and a grant to Dr. Paleolog from the British
Heart Foundation.
Ewa M. Paleolog, PhD, Mary Hunt, BSc, Michael J. Elliott,
MBBS, PhD, FRACP, Marc Feldmann, MBBS, PhD, FRCPath,
Ravinder N. Maini, MB, BChir, FRCP: Kennedy Institute of Rheumatology, London, UX, James N. Woody, MD, PhD: Centocor Inc.,
Malvern, Pennsylvania.
The clinical trial was funded by Centocor Inc., of which Dr.
Woody is an employee. Professor Feldmann is a consultant for
Centocor Inc.
Address reprint requests to Ravinder N. Maini, MB, BChir,
FRCP, Kennedy Institute of Rheumatology, 6 Bute Gardens, Hammersmith, London W6 7DW, UJS.
Submitted for publication November 15, 1995; accepted in
revised form February 8, 1996.
culature, leading to reduced migration of leukocytes into
synovial joints, and thus prolonging the therapeutic
effect of anti-TNFa in RA.
Rheumatoid arthritis (RA) is a chronic disease
characterized by peripheral synovial joint inflammation
and progressive destruction of cartilage and bone. The
lesions of RA synovitis, nearly all of which are composed
of cells derived from the blood, are an infiltrate made up
predominantly of activated CD4+ T lymphocytes and
monocyte/macrophages, associated with the generation
of a large number of cytokines and chemoattractants,
which directly and indirectly attract further leukocytes
into the synovial tissue (1). In order to extravasate into
the synovial joint, circulating cells must first adhere to
the endothelial cell layer lining the vasculature (2), a
process that is dependent on the endothelial expression
of adhesion molecules for leukocytes, including Eselectin (CD62E), intercellular adhesion molecule l
(ICAM-1) (CD54), and vascular cell adhesion molecule
1 (VCAM-1) (CD106) (3). Endothelium in RA synovium has been found to express increased levels of
E-selectin relative to either normal or osteoarthritic
synovium, predominantly on venules and capillaries (4).
In addition, immunohistochemical studies have demonstrated increased expression of VCAM-1 and ICAM-1
on RA synovial endothelial cells, although in contrast to
E-selectin, staining was also detectable on other cell
types, including synovial tissue macrophages, dendritic
cells, fibroblasts, and some lymphocytes (4-6). Expression of these adhesion molecules is dramatically upregulated on a number of cell types, including endothelium, by proinflammatory cytokines such as tumor
necrosis factor a (TNFa) and interleukin-1 (IL-1) (3),
which we and others have demonstrated to play a central
role in RA. In particular, there is considerable evidence
that TNFa, elevated levels of which are abundant in
1083
cA2 AND VASCULAR ENDOTHELIUM IN RA
synovial fluids and tissues of patients with arthritis, is of
major importance in the pathogenesis of the disease
state (for review, see refs. 1, 7, and 8).
The pivotal role of TNFa in the pathogenesis of
RA suggested that this cytokine was a potential therapeutic target, and this concept was validated in our
laboratory in the murine model of collagen-induced
arthritis, the development of which can be inhibited by
treatment with anti-TNFa after disease onset (9). In
addition, recent studies have demonstrated that
anti-TNFa treatment of rats with active arthritis markedly improved clinical scores and was paralleled by
inhibition of leukocyte accumulation in the joints (10).
We have also reported that blockade of TNFa activity in
vivo in patients with active RA by intravenous administration of monoclonal anti-TNFa antibody resulted in
marked amelioration of disease, with significant improvements observed in all clinical and laboratory parameters (11,12). Suppression of the cytokine cascade is
likely to be a cause of part of the therapeutic benefit of
anti-TNFa, since studies from our laboratory have
shown that the secretion of IL-1 and granulocytemacrophage colony-stimulating factor (GM-CSF)
(13,14), and more recently, IL-6 and IL-8 (15) by
cultured RA synovial membrane mononuclear cells was
markedly reduced by the addition of anti-TNFa antibodies. Moreover, we have also observed a rapid downregulation of circulating levels of IL-6 in RA patients
following anti-TNFa therapy (Charles P, Elliott MJ,
Davis D: unpublished observations). However, in many
patients, the clinical benefit of anti-TNFa is prolonged
(up to 26 weeks) and appears to outlast the effective
cytokine-neutralizing levels of anti-TNFa. Another possible mechanism which may account for these prolonged
effects of anti-TNFa could be diminished microvascular
endothelial activation, with consequent reduced leukocyte trafficking to the synovial joint. This hypothesis is
consistent with our observation of increased peripheral
blood lymphocyte counts in patients during the course of
anti-TNFa treatment, paralleled by reduced E-selectin
expression on capillaries and venules in synovial biopsies (16).
To further assess the significance of the decreased expression of E-selectin on synovial blood vessels, we also measured serum levels of adhesion molecules following anti-TNFa therapy. Unlike their levels in
biopsy tissues, serum levels of adhesion molecules can be
precisely quantified by enzyme-linked immunosorbent
assay (ELISA). Circulating forms of E-selectin, ICAM-1,
and VCAM-1 have been described (17), and elevated
serum concentrations of ICAM-1 and VCAM-1 have
been reported for patients with RA (18-20). In the case
of soluble E-selectin, there have been reports of both
unchanged (21) and elevated (22) serum levels in RA.
In our preliminary study, we observed that preinfusion levels of serum E-selectin, ICAM-1, and
VCAM-1 in RA were elevated relative to normal values,
but following anti-TNFa treatment, levels of serum
E-selectin, but not VCAM-1, were markedly diminished
(16). We report here that serum levels of soluble Eselectin and ICAM-l, but not VCAM-l, are significantly
reduced by anti-TNFa treatment in a time- and dosedependent manner in patients enrolled in a placebocontrolled double-blind trial of anti-TNFa chimeric
antibody. Circulating peripheral blood lymphocyte
counts were concomitantly elevated. These results have
led us to propose that the prolonged nature of the
therapeutic benefit of anti-TNFa in RA may result
partly from a major effect on the recruitment and
trafficking of blood cells to the arthritic joint.
PATIENTS AND METHODS
A total of 73 patients were recruited by 4 European
cooperatingtrial centers, in a multicenter, randomized,doubleblind, placebo-controlledclinical trial. On entry to the trial, all
patients met the American College of Rheumatology (formerly, the American Rheumatism Association) criteria for RA
(23): active RA for 2 6 months, a history of failed treatment
with at least 1 disease-modifying antirheumatic drug, and
evidence of erosive disease on radiography of the hands and
feet. Active disease was defined as the presence of 2 6 swollen
joints plus at least 3 of the following 4 secondary criteria:
duration of morning stiffness 2 4 5 minutes, 2 6 tender or
painful joints, erythrocyte sedimentation rate (ESR) 228
mm/hour, and C-reactive protein (CRP) level 220 mg/liter
(12). Due to incomplete sampling, specimens were available
for only 68 of the 73 patients.
A number of different clinical and laboratory measures
of disease activity were used for patient followup, and the
overall response to treatment was assessed by the Paulus
criteria (24). The Paulus criteria is a composite disease activity
index that defines as significant improvement in at least 4 of 6
independent variables: at least 20% improvement in the continuous variables (tender and swollen joint scores, duration of
morning stiffness, and ESR) and at least a grade 2 improvement in the patient’s and the observer’s assessments of disease
severity.
The antibody to TNFa (cA2) was a humadmurine
chimeric monoclonal antibody of IgGl K isotype, with specificity for recombinant and natural human TNFa, and was produced by Centocor Inc. (25). Study patients received a single
intravenous infusion of either placebo (n = 24) or cA2 (n =
44) at a concentration of either 1 mgkg (n = 23) or 10 mg/kg
(n = 21). Serum samples, taken between 8:OO AM and 12:OO PM
to reduce the influence of diurnal variations in serum adhesion
molecule levels (26), were obtained prior to infusion and at
PALEOLOG ET AL
1084
Table 1. Decreases in serum E-selectin and ICAM-I, hut not VCAM-1, levels in rheumatoid arthritis patients after infusion of monoclonal
anti-TNFa antibody cA2*
cA2
Parameter,
week
E-selectin
Week 0
Week 2
Week 4
ICAM-1
Week 0
Week 2
Week 4
VCAM-1
Week 0
Week 2
Week 4
Placebo (n
=
24)
1 m&g (n
=
10 m&g (n
23)
=
21)
SD
Range
Median
SD
Range
Median
SD
Range
63.6
69.7
60.0
49.1
59.0
47.1
15.6-183.7
15.3-274.4
16.5-210.6
58.3
53.3
54.1
27.5
25.9
26.6
22.0-113.9
15.1-108.1
15.5-1 18.1
58.4
41.2
37.4
51.1
43.2
40.6
24.4-185.3
20.3-169.0
20.5-192.6
437.7
438.0
454.8
209.8
229.9
259.3
193.7-1,045.8
200.8 -1 ,O 18.8
207.3-1,134.0
421.8
353.2
398.6
166.5
148.0
182.4
202.6-795.6
174.9- 843.0
253.2-1,040.8
472.3
355.5
305.9
167.3
159.4
154.3
266.5-1.022.4
204.6 -929.0
197.0-730.9
830.5
864.4
872.3
302.4
367.4
329.4
404.4-1,685.0
478.3-1,855.0
478.3-1,755.8
802.5
929.3
990.1
298.4
332.9
358.4
263.2-1,53 1.5
387.2-1,555.2
427.3-1,773.5
893.0
780.5
746.2
254.2
302.5
324.4
517.8-1,419.1
452.0-1,453.7
563.3-1,637.9
Median
* Values are in ng/ml at preinfusion (week 0) and either 2 or 4 weeks postinfusion. ICAM-1
vascular cell adhesion molecule 1; anti-TNFa = anti-tumor necrosis factor a.
weekly intervals thereafter, up to a total of 4 weeks after
infusion.
Serum E-selectin, ICAM-1, and VCAM-1 were assayed in a blinded manner by ELISA (R&D Systems, Abingdon, Oxon, UK) according to the manufacturer’s instructions,
and adhesion molecule concentrations were calculated by
reference to a standard curve, detection limits for which were
0.5 ng/ml, 2.0 ng/ml, and 5.0 ng/ml for soluble E-selectin,
ICAM-1, and VCAM-1, respectively. All samples from a single
patient were assayed on a single ELISA plate, and a control
sample of serum of known adhesion molecule concentrations
was assayed on every plate. Addition of diluent containing up
to 100 pghl of cA2 was not found to significantly affect the
results obtained, indicating that any residual serum cA2 was
not interfering with the soluble adhesion molecule assays. The
intraassay variation was found not to exceed lo%, and the
interassay variation was 15%. Mean serum levels in normal
individuals, as determined by the manufacturers, are 46 ng/ml
(range 29-63), 210 ng/ml (range 115-306), and 553 ng/ml
(range 395-714) for E-selectin, ICAM-1, and VCAM-1, respectively (17,27). Leukocyte counts for all 73 patients were
performed on fresh blood samples collected into tubes containing EDTA. A n automated Coulter cell counter was used
for total leukocyte and some differential cell counts, with the
remaining differentials performed manually on blood smears.
Results are expressed as the medians ? SD.
There were no study dropouts. All points therefore
represent a total of 24, 23, or 21 patients for the placebo-, the
1 mg/kg of cA2-, and the 10 mg/kg of cA2-treated groups,
respectively. Where values are expressed relative to preinfusion levels, calculations were separately performed for each
patient, prior to calculating the median percentages of change
after treatment.
Statistical analyses were performed by Wilcoxon sign
rank test for comparisons within patient groups, using individual patient data sets as input variables. For comparisons
between treatment groups at different time points, data were
analyzed by Mann-Whitney U test, using both individual
=
intercellular adhesion molecule 1; VCAM-1
=
patient data and percentages of change relative to the preinfusion values as the response variables. Significance values for
comparisons between multiple groups were adjusted using the
Bonferroni correction.
RESULTS
Changes in circulating soluble adhesion molecule levels after cA2 treatment. Initial serum levels of
E-selectin were elevated in 44 of the 68 patients (relative
to mean serum levels in normal individuals, equivalent
to 46 t 17 ng/ml, median 44 ng/ml, range 24-80 ng/ml
[ref. 17 and Paleolog EM, Hunt M: unpublished observations]). Median preinfusion E-selectin levels were 64
ng/ml for the placebo group, 58 ng/ml for the 1 mgkg
cA2-treated group, and 58 ng/ml for the 10 mg/kg
cA2-treated group (Table 1). However, between-group
differences in preinfusion levels of E-selectin were not
statistically significant by Mann-Whitney U test (P >
0.50 for placebo versus 1 mdkg cA2 group and P > 0.95
for placebo versus 10 mgkg cA2 group), and the values
were normalized relative to preinfusion serum concentrations of E-selectin for each patient (Figure 1). All
subsequent analyses by Mann-Whitney U test were
performed using the percentage of change relative to the
median preinfusion value as a response variable.
Circulating levels of E-selectin were unchanged
in the placebo-treated patients over the course of the 4
weeks for which samples were assayed. In contrast, in
both of the anti-TNFa-treated groups, serum E-selectin
levels decreased 1 week postinfusion. For example, 1
week after cA2, median E-selectin levels decreased from
cA2 AND VASCULAR ENDOTHELIUM IN RA
1101
t
.
lmg/kgcA2
B 1
rn
70
I
8
0
1
2
3
4
Weeks after infusion
Figure 1. Decreases in serum E-selectin levels after anti-tumor necrosis factor a (anti-TNFa [cA2]) infusion. Values are expressed
relative to median preinfusion values (100%) in 24 placebo-treated, 23
1 mgkg of cA2-treated, and 21 10 mfig of cA2-treated patients.
Significance was determined versus preinfusion E-selectin levels by
Wilcoxon sign rank test: * = P 5 0.05, ** = P 5 0.01, and *** = P 5
0.001; and versus change in placebo-treated group at different time
points by Mann-Whitney U test: t = P 5 0.05, tt = P 5 0.01, and
ttt
=
P
5
0.001.
58.3 ng/ml to 50.9 ng/ml in the 1 mgkg cA2-treated
group (P < 0.05 versus change in placebo-treated
group), and from 58.4 ng/ml to 43.5 ng/ml for the 10
mg/kg cA2-treated group (P < 0.001 versus change in
placebo-treated group). The peak effect was observed at
week 2 postinfusion for the 1 mgkg cA2-treated individuals (decrease of 22%; P < 0.01 versus change in
placebo group), and at week 1 for the 10 mgkg cA2treated group (decrease of 26%; P < 0.001 versus
change in placebo group) (Figure 1).
Initial serum levels of ICAM-1 (2210 ng/ml) (17)
were elevated in 64 of the 68 patients. Median preinfusion soluble ICAM-1 levels were comparable among the
3 groups: 438 ng/ml in the placebo group, 422 ng/ml in
the 1 m a g cA2 group, and 472 ng/ml in the 10 mgkg
cA2 group (Table 1). All values were normalized as
described above, and analyses by Mann-Whitney U test
were performed using the percentage of change relative
to the median preinfusion value as a response variable.
In the placebo-treated patients, soluble ICAM-1
levels did not change substantially at any time point,
whereas these levels decreased significantly (P < 0.01
versus change in placebo group) in both cA2-treated
groups 1 week after infusion. For the group treated with
1 mgkg of cA2, soluble ICAM-1 levels decreased maximally (by 19%) at week 2 (P < 0.01 versus week 0 and
1085
versus change in placebo group). However, unlike Eselectin, which was still decreased at this time point in
the low-dose cA2 group, the serum ICAM-1 concentrations returned to preinfusion values by week 3. Circulating ICAM-1 levels in the 10 mg/kg cA2-treated patients
were diminished at weeks 1-4, with the maximum
decrease (29%) at week 4 (P < 0.001 versus week 0 and
versus change in placebo group) (Figure 2).
In addition, soluble E-selectin levels were measured in a limited number of patients on days l and 3
after infusion (7 patientshreatment group). On day 1,
there were modest (110-15%) decreases in circulating
levels of both ICAM-1 and E-selectin in both the
low-dose and the high-dose treatment groups. On day 3,
serum E-selectin and ICAM-1 decreased by up to 13%
in the 1 m a g cA2-treated group, whereas in patients
who received 10 mgkg cA2, the changes were 23% and
17% for E-selectin and ICAM-1, respectively (data not
shown). However, because of the small numbers of
patients in each group, these differences were not found
to differ significantly from the changes in the placebotreated patients.
Circulating VCAM-1 levels were elevated (2553
ng/ml) (18) in 61 of the 68 patients, but were not
significantly affected by anti-TNFa (Table 1). There was
no significant difference in VCAM-1 levels compared
I"
0
1
2
3
4
Weeks after infusion
Figure 2. Decreases in serum intercellular adhesion molecule 1
(ICAM-1)
levels after anti-tumor necrosis factor a (anti-TNFa [cA2])
infusion. Values are expressed relative to median preinfusion values
(100%) in 24 placebo-treated, 23 1 mg/kg of cA2-treated, and 21 10
mg/kg of cA2-treated patients. Significance was determined versus
preinfusion ICAM-1levels by Wilcoxon sign rank test: * = P 5 0.05,
** = P 5 0.01, and *** = P zs 0.001; and versus change in
placebo-treated group at different time points by Mann-Whitney U
test: tt = P 5 0.01, and ttt = P 5 0.001.
PALEOLOG ET AL
Table 2. Peripheral blood lymphocyte and neutrophil counts in rheumatoid arthritis patients after infusion of monoclonal anti-TNFa antibody cA2'
cA2
Parameter,
day
Lymphocytes
Day 0
Day 1
Day 3
Day 7
Day 28
Neutrophls
Day 0
Day 1
Day 3
Day 7
Day 28
Placebo (n
=
1 mgkg (n = 23)
24)
Median
SD
Range
1.50
1.60
1.50
1.53
1.30
0.90
0.93
0.69
0.77
0.85
7.50
5.60
7.20
6.95
7.65
3.14
2.70
3.29
2.83
2.70
10 mg/kg (n
Median
SD
Range
0.50-4.00
0.90-4.13
0.49-3.20
0.60-3.62
0.55-4.16
1.30
2.20
2.00
1.98
1.56
0.77
1.14
0.67
0.81
0.81
2.80-14.90
3.80-13.30
3.80-15.00
4.40-14.20
4.10-14.50
7.40
4.65
4.90
6.15
6.45
2.91
2.11
3.67
2.69
3.87
21)
Median
SD
Range
0.81-3.70
1.30-6.30
1.30-4.10
0.61-3.90
0.90-4.90
1.60
2.40
2.25
2.40
2.00
0.45
0.54
0.61
0.60
0.52
0.90-2.80
1.30-3.30
1.50-3.50
1.50-3.60
1.10-3.00
3.80-15.00
2.50-10.20
2.70-17.00
2.30-14.00
2.40-21.00
6.10
3.90
3.95
4.60
4.70
1.76
1.45
2.27
2.56
2.02
3.20-1 1.00
1.70 -7.20
1.90-1 1.00
2.40-12.00
1.40-8.50
* Values are XlO9//liter,at preinfusion (day 0) and on day 1, 3, 7, or 28 postinfusion. Anti-TNFa
with preinfusion or with the placebo-treated group at
any time point.
Changes in circulating leukocyte counts after
cA2 treatment. Pretreatment lymphocyte counts were at
the lower extreme of the normal range for all 3 patient
groups, and remained stable in the placebo-treated
group. In contrast, cA2-treated patients exhibited a
rapid rise in lymphocyte counts following anti-TNFa
infusion, with the earliest increase on day 1 after infusion (Table 2 and Figure 3A). In the 1 m a g cA2treated group, lymphocyte counts on days 1 and 3 were
significantly increased (P 5 0.01 versus preinfusion), but
subsequently showed a gradual decline toward the initial
levels (no significant difference versus preinfusion levels
at weeks 2-4). Lymphocyte counts peaked on day 3 for
the patients who received 10 mg/kg of cA2 (maximum
increase 53%; P < 0.001 versus preinfusion values and
versus change in the placebo-treated group). The elevation in circulating lymphocyte levels persisted over the
course of the study in the high-dose cA2 group (median
increase at week 4 -36%).
Circulating neutrophils showed a more complex
pattern of change, with initial counts at the higher
extreme of the normal range, followed by a drop in day
1 values in all 3 groups (P 5 0.05 for all groups versus
preinfusion) (Table 2 and Figure 3B). However, the
extent of the decrease in day 1 neutrophil counts was
greater in cA2-treated patients (average decrease 36%)
than in patients who received placebo (average decrease
18%). Neutrophil counts subsequently recovered on day
3 in the placebo-treated patients. A partial recovery was
also observed in the cA2-treated patients, but the neu-
=
=
anti-tumor necrosis factor a.
trophil counts did not revert to the initial values over
the course of the study for either the low-dose or the
high-dose groups. Counts were consistently lower in
the high-dose cA2 group (decrease 26% at week 4)
relative to the low-dose group (decrease 17% at week 4)
(Figure 3B), and even 4 weeks after infusion of 10 mg/kg
of cA2, the counts were still significantly below pretreatment levels (P < 0.01).
Peripheral blood monocyte counts, in contrast,
showed a trend toward lower values in cA2-treated
patients, particularly on days 1 and 3 posttreatment. For
example, on day 3 after cA2 infusion, the median
decrease in peripheral blood monocytes was 20% and
29% for the low- and high-dose groups, respectively
( P < 0.01 versus change in placebo group). However, at
subsequent time points, the monocyte counts returned
to preinfusion levels (data not shown).
Correlation of changes in the levels of soluble
adhesion molecules and leukocyte counts with changes
in clinical parameters. We compared the changes in
serum levels of E-selectin, ICAM-1, and peripheral
blood lymphocytes with the degree of disease amelioration in individual patients 4 weeks after infusion of cA2,
since the designated end-point of the study was the
achievement of a 20% response on the Paulus criteria 4
weeks after infusion of cA2 (12). Anti-TNFa-treated
patients were subdivided according to the absence and
presence of the 20% Paulus response at this time point.
Eleven of the 23 patients (48%) who received 1 mg/kg of
cA2 and 17 of the 21 patients (81%) who received 10
mgkg of cA2 had achieved the 20% Paulus response at
week 4. Among the two cA2 treatment groups com-
1087
cA2 AND VASCULAR ENDOTHELIUM IN RA
preinfusion values), whereas in the group of 28 patients
in whom a 20% Paulus response was observed at this
time, only 3 patients failed to show a reduction in
E-selectin (median E-selectin levels 77%) ( P < 0.01
versus absence of 20% Paulus response, by MannWhitney U test).
Similarly, median ICAM-1 levels were 99% in
those with an absence of the 20% Paulus response and
77% in those with the 20% Paulus response relative to
preinfusion values ( P < 0.01 presence versus absence of
20% Paulus response, by Mann-Whitney U test) (Figure
4B). Comparable analyses were performed for lymphocyte counts (Figure 4C) at week 4 after infusion of
anti-TNFa. For the individuals in whom the 20% Paulus
criteria were not met, there was no change in the median
number of circulating lymphocytes versus week 0 (median increase 0% relative to week 0; 9 of 16 patients
unchanged versus preinfusion), as compared with an
increase of 29% in the patients in whom a 220% Paulus
response was observed (4 of 28 patients unchanged
versus preinfusion; P < 0.05 presence versus absence of
20% Paulus response).
DISCUSSION
Figure 3. Changes in circulating leukocyte and neutrophil counts
after anti-tumor necrosis factor a (anti-TNFa [cA2]) infusion. Values
are the median counts in 24 placebo-treated, 23 1 mg/kg of cA2treated, and 21 10 mgkg of cA2-treated patients. Circulating numbers
of lymphocytes (A) and neutrophils (B) in patients with active rheumatoid arthritis were measured before and after treatment. Significance was determined versus preinfusion levels by Wilcoxon sign rank
test: * = P 5 0.05, ** = P 5 0.01, and *** = P 5 0.001; and versus the
placebo-treated group at different time points by Mann-Whitney U
test: t = P 5 0.05, tt = P zs 0.01, and ttt = P 5 0.001.
bined, the 20% Paulus response was absent in 16 patients and present in 28 patients.
Figure 4A illustrates the change in serum Eselectin concentrations relative to preinfusion values as a
function of the presence or absence of a 20% Paulus
response 4 weeks after infusion of anti-TNFa (1 or 10
mg/kg). It is clear that in the group of 16 patients who
did not meet the 20% Paulus criteria at this time, 7
individuals also did not exhibit any decrease in Eselectin levels (median E-selectin levels 95% relative to
In this study, we have demonstrated a rapid and
statistically significant decrease in the circulating levels
of E-selectin and ICAM-1, but not VCAM-1, in RA
patients following infusion of anti-TNFa antibody cA2.
These findings confirm and extend our previous results
(16). Both the extent and duration of the decrease in
serum E-selectin and ICAM-1 were dependent on the
dose of cA2. The time- and dose-dependence of the
reductions in circulating E-selectin and ICAM-1 observed in this study was similar to the pattern of changes
in CRP levels and ESR previously demonstrated after
cA2 treatment (12). The earliest observable changes in
serum E-selectin and ICAM-1 occurred on days 1 and 3
after anti-TNFa, although the decreases did not achieve
statistical significance until 1 week after anti-TNFa
infusion. In the patients who received 1 mg/kg of cA2,
the decrease in serum adhesion molecules was transient,
returning to preinfusion levels by week 3 for ICAM-1
and by week 4 for E-selectin, whereas the diminution in
the levels of circulating adhesion molecules for the
patients treated with high-dose cA2 persisted even at 4
weeks, mimicking the longer-term clinical effect of the
higher dose of anti-TNFa (12). In contrast to the
decreases in serum levels of E-selectin and ICAM-I,
numbers of circulating lymphocytes increased rapidly
PALEOLOG ET AL
1088
B. Serum ICAM-1
A. Serum E-selectin
I
...........
...........I,
I
.
0
.............................
0
rl,
8
.ii:
I
v0
I
i
0
0.
0
)
Absent
Present
20% Paulus Response
0
I
*
rr
0
I
pcO.05
0
0
p<o.m
'
....
w
pcO.01
30
I
-:
0
:
0
l
-"p
C. Lymphocyte counts
1
0
I ;
4
Absent
Present
20% Paulus Response
Absent
Present
20% Paulus Response
Figure 4. Correlation of changes in levels of soluble adhesion molecules and circulating lymphocytes with clinical benefit. Correlation between the
presence and absence of a 20% Paulus response, a composite index of disease activity measurement (25), at 4 weeks after infusion of anti-tumor
necrosis factor a antibody cA2 (1 or 10 mg/kg; n = 23 and n = 21, respectively), and either A, the percentage decrease in serum E-selectin levels,
B,the percentage decrease in serum intercellular adhesion molecule 1 (ICAM-1) levels, or C, the percentage increase in peripheral blood lymphocyte
counts. Box plots represent median values; hinges enclose the 25th and 75th percentiles. n = 16 for absence of 20% Paulus response (12 of 23
patients taking 1 mg/kg and 4 of 21 patients taking 10 mgkg of cA2) and n = 28 for presence of 20% Paulus response (11 of 23 patients taking 1
mg/kg and 17 of 21 patients taking 10 mgkg of cA2). Statistical analyses were performed by Mann-Whitney U test.
after infusion of anti-TNFa, although neutrophil and
monocyte counts were concomitantly reduced.
Although the changes in serum E-selectin levels
appeared to be relatively modest (maximum decrease
26%), it is noteworthy that after 4 weeks, median serum
E-selectin levels in patients who received 10 mg/kg of
cA2 were 37 ng/ml (range 21-193), well within the range
of values reported for normal individuals (mean serum
levels 46 ng/ml, range 24-80 ng/ml [ref. 17 and Paleolog
EM, Hunt M: unpublished observations]). Serum
ICAM-1 levels after 10 mg/kg of cA2 (median 306 ng/ml
at week 4) were close to the upper limit of the range
reported for normal individuals (mean 210 ng/ml, range
115-306 ng/ml [17]). Our results provide the first detailed account of a reduction in serum E-selectin and
ICAM-1 as part of the response to any biologic therapy
in human disease, and support our previously published
preliminary data (16). There is, to date, only limited
evidence showing the effects of antirheumatic drugs on
endothelial cell biology. For example, decreased synovial
expression of E-selectin has been described following
gold sodium thiomalate treatment (28), suggesting that
its antiinflammatory activity may be partly due to an
effect on endothelial cell adhesion molecules. However,
these studies are complicated by the requirement for
serial synovial biopsy samples, and the recent identification of soluble forms of adhesion molecules has allowed
a more detailed and quantitative analysis of the effect of
therapy on leukocyte-endothelial cell interactions. Sulfasalazine has been demonstrated to significantly decrease levels of serum E-selectin after 3 months of
therapy (29), although the changes (average decrease
11%) were more modest and the time course much
slower than that of anti-TNFa.
The reason for the discrepancy between the effect
of anti-TNFa therapy on serum E-selectin and ICAM-1
(in terms of differing time courses and extent of the
changes), and the absence of an effect on serum
VCAM-1, is unclear, but may possibly be related to
different sources of these soluble adhesion molecules o r
to different in vivo responsiveness to cytokines, for
example, the selective induction of VCAM-1 by IL-4, or
the very rapid (within 4 hours) and transient induction of
E-selectin expression, as compared with the more prolonged ( 2 2 4 hours) and sustained induction of ICAM-1
and VCAM-1 (3). E-selectin is expressed selectively on
endothelium, in contrast to ICAM-1 and VCAM-1,
potential in vivo sources of which include cells other
cA2 AND VASCULAR ENDOTHELIUM IN RA
than endothelium (such as lymphocytes, fibroblasts,
macrophages, and dendritic cells). Serum E-selectin
must therefore, by implication, be endothelial cellderived, and hence alterations in the serum levels of this
adhesion molecule in disease states may reflect changes
in the endothelial cell phenotype in vivo. Interestingly,
release of E-selectin, ICAM-I, and VCAM-1 from cultured human vascular endothelial cells has been shown
to vary in proportion of the degree of surface expression
in response to cytokines such as TNFa, which is well
established as an important inducer of the expression of
these adhesion molecules in vitro (30,31). However, the
effect of TNFa on the release of ICAM-1 and VCAM-1
from other cell types (such as fibroblasts) has not been
studied. The possible in vivo roles of soluble adhesion
molecules are, at present, unclear, although immobilized
soluble VCAM-1 and E-selectin have been found to
support the binding of human leukocytes and leukocyte
cell lines (20,32,33). More recently, soluble E-selectin
and VCAM-1 were demonstrated to promote angiogenesis in rat cornea and to induce chemotaxis of human
endothelial cells (34), which may be of particular relevance in RA, since angiogenesis contributes significantly
to the pathogenesis of the disease state (35).
The magnitudes of the decreases in serum levels
of E-selectin and ICAM-1 and the increases in the
numbers of circulating lymphocytes in the cA2-treated
patients were signiwntly higher ( P 5 0.05) in patients
who showed substantial improvements in clinical parameters (220% Paulus response at week 4 postinfusion)
compared with those who received anti-TNFa but failed
to show any disease amelioration at the end of the study.
The reciprocal changes in serum adhesion molecule
levels and circulating lymphocyte counts suggest that
decreased endothelial cell expression of adhesion molecules may contrib
the rise in peripheral blood
lymphocyte counts
hypothesis is supported by our
data demonstratin
rease in the mean proportion
E-selectin in paired synovial
of blood vessels e
biopsies taken before and after anti-TNFa treatment,
from 18.6% (mean of at least 3 samples per patient, SD
11.1%) to 9.1% (SD 5.6%), together with a marked
reduction in the degree of cellular infiltration, the
thickness of the lining layer, and the number of lymphoid aggregates in the RA membrane after cA2 (36).
Suppression of arthritis as a result of in vivo blockade of
leukocyte adhesion has been demonstrated both in
animal models of disease (37-39) and in humans with
RA (40). Moreover, anti-TNFa treatment of rats with
active arthritis resulted in marked amelioration of disease, paralleled by inhibition of leukocyte accumulation
1089
in the joints (lo), further emphasizing the importance of
leukocyte recruitment in the pathogenesis of arthritis,
and supporting our hypothesis that the therapeutic
benefit of cA2 may be partly achieved via downregulation of leukocyte trafficking to the joints.
The same mechanism cannot, however, explain
the decreases in circulating neutrophils (and, to a lesser
extent, monocytes) following TNFa blockade. The initial
decreases in neutrophil counts on day 1in all groups may
have their origin in factors common to all patients
groups, such as bed rest or hemodilution. Subsequently,
the neutrophil counts in the placebo-treated subjects
rapidly (by day 3) reversed to preinfusion levels, but
remained decreased in the anti-TNFa-treated patients.
Interestingly, synovial fluid neutrophil counts also declined in 3 of 4 patients in whom joint aspiration was
performed after infusion of anti-TNFa (Elliott MJ,
Taylor PC: unpublished observations), which suggests
that inhibition of the production of hemopoietic growth
factors, such as GM-CSF (14) or granulocyte-CSF, may
account for our findings of decreased circulating levels
of these relatively short-lived cells following anti-TNFa
treatment.
Our observations may offer important insights
into the mechanism of action of anti-TNFa in ameliorating the inflammation of RA. In particular, we have
observed a greater decrease in serum E-selectin and
ICAM-1 levels, paralleled by a significantly higher increase in peripheral lymphocyte counts, in cA2-treated
patients who had a marked clinical response to
anti-TNFa, thus strengthening the clinical relevance of
the laboratory measurements and validating the in vivo
relevance of the in vitro observations of induction of
endothelial cell adhesion molecule expression in response to TNFa.
Reduced recruitment of inflammatory cells to the
synovium would appear, at least in part, to be a consequence of decreased adhesion molecule expression by
the synovial microvascular endothelium. Such an effect
may be either a direct result of the in vivo neutralization
of TNFa by cA2, and hence diminished endothelial
activation, or a secondary effect, due to the downmodulation of monocyte/macrophage responses, such as
the production of proinflammatory cytokines, which can
activate endothelium. However, the rapid time course of
the alterations in levels of serum adhesion molecules
and lymphocyte counts (earliest observed decrease 1-3
days) is strongly suggestive of a primary effect of
anti-TNFa on endothelial cell activation, and thus on
leukocyte trafficking. Nevertheless, leukocyte trafficking
requires not only the expression of adhesion molecules
PALEOLOG ET AL
1090
by vascular endothelium, but also a second signal, which
is provided by chemotactic factors such as chemokines.
In this context, it is of interest that anti-TNFa blocks RA
synovial cell production of the chemokine IL-8 (15), and
there is preliminary evidence from our laboratory of
decreased serum IL-8 levels in RA patients after
anti-TNFa infusion, indicating that anti-TNFa may
down-regulate both arms of the leukocyte trafficking
system (Charles P, Hogg A, Paleolog EM: unpublished
observations).
Taken together, our results suggest that downregulation of lymphocyte trafficking to the arthritic joint
after cA2 infusion is an important component of the
improvement observed in RA patients as a consequence
of anti-TNFa therapy, and may account for the prolonged duration of clinical benefit of anti-TNFa in RA.
The hypothesis that anti-TNFa reduces trafficking to the
joints can only be formally verified by migration studies
using labeled cells, and such studies are being planned.
ACKNOWLEDGMENTS
We thank the European centers who have collaborated
in the clinical trial, in particular, Dr. F. C. Breedveld and Dr.
J. D. Macfarlane (Department of Rheumatology, University
Hospital, Leiden, The Netherlands), Dr. J. S. Smolen and Dr.
B. Leeb (University Clinic for Internal Medicine 111, Vienna,
Austria), and Dr. J. R. Kalden and Dr. C. Antoni (Institute of
Clinical Immunology and Rheumatology, Erlangen, Germany). We also wish to thank the staff of Centocor Inc.: Ms.
Kim de Woody for help with statistical analysis and Dr. Hanny
Bijl for coordinating the collection of sera and clinical data.
REFERENCES
1. Feldmann M, Brennan FM, Maini RN: Role of cytokines in
rheumatoid arthritis. Annu Rev Immunol 14:397-440, 1995
2. Cronstein BN: Adhesion molecules in the pathogenesis of rheumatoid arthritis. Curr Opin Rheumatol 6:300-304, 1994
3. Carlos TM, Harlan JM: Leukocyte-endothelial adhesion molecules. Blood 84:2068-2101, 1994
4. Koch AE, Burrows JC, Haines GK, Carlos TM, Harlan JM,
Leibovich SJ: Immunolocalization of endothelial and leukocyte
adhesion molecules in human rheumatoid and osteoarthritic synovial tissues. Lab Invest 64:313-620, 1991
5. Johnson BA, Haines GK, Harlow LA, Koch AE: Adhesion
molecule expression in human synovial tissue. Arthritis Rheum
36~137-146, 1993
6. Szekanecz Z, Haines GK, Lin TR, Harlow LA, Goerdt S, Rayan
G, Koch AE: Differential distribution of intercellular adhesion
molecules (ICAM-1, ICAM-2, and ICAM-3) and the MS-1 antigen
in normal and diseased human synovia: their possible pathogenetic
and clinical significance in rheumatoid arthritis. Arthritis Rheum
37~221-231, 1994
7. Brennan FM, Maini RN, Feldmann M: TNFa-a pivotal role in
rheumatoid arthritis? Br J Rheumatol 31:293-298, 1992
8. Arend WP, Dayer J-M: Inhibition of the production and effects of
interleukin-1 and tumor necrosis factor a in rheumatoid arthritis.
Arthritis Rheum 38:151-160, 1995
9. Williams RO, Feldmann M, Maini RN: Anti-tumor necrosis factor
ameliorates joint disease in murine collagen-induced arthritis.
Proc Natl Acad Sci U S A 89:9784-9788, 1992
0. Issekutz AC, Meager A, Otterness I, Issekutz TB: The role of
tumor necrosis factor-a and IL-1 in polymorphonuclear leukocyte
and T lymphocyte recruitment to joint inflammation in adjuvant
arthritis. Clin Exp Immunol 97:26-32, 1994
1. Elliott MJ, Maini RN, Feldmann M, Long-Fox A, Charles P,
Katsikis P, Brennan FM, Walker J, Bijl H. Ghrayeb J, Woody JN:
Treatment of rheumatoid arthritis with chimeric monoclonal
antibodies to tumor necrosis factor a. Arthritis Rheum 36: 16811690, 1993
12. Elliott MJ, Maini RN, Feldmann M, Kalden JR, Antoni C, Smolen
JS, Leeb B, Breedveld FC, Macfarlane JD, Bijl H, Woody JN:
Randomised double-blind comparison of chimeric monoclonal
antibody to tumor necrosis factor a (cA2) versus placebo in
rheumatoid arthritis. Lancet 344:1105-1110, 1994
13. Brennan FM, Chantry D, Jackson A, Maini R,. Feldmann M:
Inhibitory effect of TNFa antibodies on synovial cell interleukin-1
production in rheumatoid arthritis. Lancet 2:244-247, 1989
14. Haworth C, Brennan FM, Chantry D, Turner M, Foxwell B, Maini
RN, Feldmann M: Expression of granulocyte-macrophage colonystimulating factor in rheumatoid arthritis: regulation by tumour
necrosis factor-a. Eur J Immunol 21:2575-2579, 1991
15. Butler DM, Maini RN, Feldmann M, Brennan FM: Modulation of
proinflammatory cytokine release in rheumatoid synovial membrane cell cultures with an anti-TNFa monoclonal: comparison
with blockade of IL-1 using the recombinant IL-1 receptor antagonist. Eur Cytokine Netw 6:225-230, 1995
16. Maini RN, Elliott MJ, Brennan FM, Williams RO, Chu CQ,
Paleolog E, Charles PJ, Taylor PC, Feldmann M: Monoclonal
anti-TNFa antibody as a probe of pathogenesis and therapy of
rheumatoid disease. Immunol Rev 144:195-223, 1995
17. Gearing AJH, Newman W: Circulating adhesion molecules in
disease. Immunol Today 14506-512, 1993
18. Cush JJ, Rothlein R, Lindsley HB, Mainolfi EA, Lipsky PE:
Increased levels of circulating intercellular adhesion molecule 1 in
the sera of patients with rheumatoid arthritis. Arthritis Rheum
36: 1098-1 102, 1993
19. Mason JC, Kapahi P, Haskard DO: Detection of increased levels
of circulating intercellular adhesion molecule 1 in some patients
with rheumatoid arthritis but not in patients with systemic lupus
erythematosus: lack of correlation with levels of circulating vascular cell adhesion molecule 1. Arthritis Rheum 36519-527, 1993
20. Wellicome SM, Kapahi P, Mason JC, Lebranchu Y,Yarwood H,
Haskard DO: Detection of a circulating form of vascular cell
adhesion molecule-1: raised levels in rheumatoid arthritis and
systemic lupus erythematosus. Clin Exp Immunol 92:412-418,
1993
21. Koch AE, Turkiewicz W, Harlow LA, Pope RM: Soluble Eselectin in arthritis. Clin Immunol Immunopathol 69:29-35, 1993
22. Blann AD, McCollum CN, Steiner M, Jayson MIV: Circulating
adhesion molecules in inflammatory and atherosclerotic vascular
disease. Immunol Today 16:251-252, 1995
23. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS,
Medsger TA Jr, Wilder RL, Hunder GG: The American Rheumatism Association 1987 revised criteria for the classification of
rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988
24. Paulus HE, Egger MJ, Ward JR, Williams HJ, the Cooperative
Systematic Studies of Rheumatic Diseases Group: Analysis of
improvement in individual rheumatoid arthritis patients treated
with disease-modifying antirheumatic drugs, based on the findings
in patients treated with placebo. Arthritis Rheum 33:477-484,
1990
cA2 AND VASCULAR ENDOTHELIUM IN RA
25. Knight DM, Trinh H, Le J, Siege1 S, Shealy D, McDonough M,
Scallon B, Moore MA, Vilcek J, Daddona P, Ghrayeb J: Construction and initial characterization of a mouse-human chimeric
anti-TNF antibody. Mol Immunol 30:1443-1453, 1993
26. Maple C? Kirk G, McLaren M, Belch JJF: A diurnal variation
exists in soluble E-selectin, P-selectin and ICAM-1 levels in
healthy volunteers (abstract). Br J Rheumatol 34 (suppl 1):87.
1995
27. Gearing AJ, Hemingway I, Pigott R, Hughes J, Rees AJ, Cashman
SJ: Soluble forms of vascular adhesion molecules. E-selectin,
ICAM-I and VCAM-1: pathological significance. Ann N Y Acad
Sci 667:324-331. 1992
28. Corkill MM, Kirkharn BW: Haskard DO, Barbatis C, Gibson T,
Panayi GS: Gold treatment of rheumatoid arthritis decreases
synovial expression of the endothelial leukocyte adhesion receptor
ELAM-1. J Rheumatol 18:1453-1460, 1991
29. Maple C. Kirk G, Veale D, Belch J J F Reduction in cell adhesion
molecule levels following sulphasalazine in patients with rheumatoid arthritis (abstract). Br J Rheumatol 34 (suppl 1):98, 1995
30. Leeuwenberg JF, Smeets EF, Neefjes JJ, Shaffer MA, Cinek T,
Jeunhomme TM, Ahern TJ. Buurrnan WA: E-selectin and intercellular adhesion molecule-1 are released by activated human
endothelial cells in vitro. Immunology 77543-549, 1992
31. Pigott R, Dillon LP, Hemingway IH, Gearing AJ: Soluble forms of
E-selectin, ICAM-1 and VCAM-1 are present in the supernatants
of cytokine activated cultured endothelial cells. Biochem Biophys
Res Cornniun 187584-589, 1992
32. Lobb R, Chi-Rosso G, Leone D, Rosa M, Newman B, Luhowskyj
S, Osborn L, Schitier S, Benjamin C. Dougas I, Hession C, Chow
P: Expression and functional characterization of a soluble form of
vascular cell adhesion molecule-1. Biochem Biophys Res Commun
17811498-1504, 1991
1091
33. Lobb RR, Chi-Rosso G, Leone DR, Rosa MD, Bixler S, Newman
BM. Luhowskyj S, Benjamin CD, Dougas IG, Goelz SE, Hession
C . Chow EP: Expression and functional characterization of a
soluble form of endothelial-leukocyte adhesion molecule-I. J
Imrnunol 147:124-129, 1991
34. Koch AE, Halloran MM, Haskell CJ, Shah MR, Polverini PJ:
Angiogenesis mediated by soluble forms of E-selectin and vascular
cell adhesion molecule-I. Nature 376517-519, 1995
35. Colville-Nash PR, Scott DL: Angiogenesis and rheumatoid arthritis: pathogenic and therapeutic implications. Ann Rheum Dis
51:919-925, 1992
36. Paleolog EM, Taylor PC, Hunt M, Tak PP. Elliott MJ. Feldmann
M, Breedveld FC, Maini RN: Treatment of rheumatoid arthritis
with antibody to TNFa decreases expression and shedding of
E-selectin (abstract). Arthritis Rheum 38 (suppl 9):S279, 1995
37. Kakimoto K, Nakamura T, Ishii K, Takashi T, Iigou H, Yagita H,
Okumura K. Onoue K: The effect of anti-adhesion molecule
antibody on the development of collagen-induced arthritis. Cell
Immunol 142:326-337. 1992
38. Issekutz AC, Issekutz TB: A major portion of polymorphonuclear
leukocyte and T lymphocyte migration to arthritic joints in the rat
is via LFA-l/MAC-1-independent mechanisms. Clin Imrnunol
Immunopathol 67257-263, 1993
39. Wahl SM, Allen JB, Hines KL, Imamichi T, Wahl AM, Furcht LT,
McCarthy JB: Synthetic fibronectin peptides suppress arthritis in
rats by interrupting leukocyte adhesion and recruitment. J Clin
Invest 94:65S-662, 1994
40. Kavanaugh AF,Davis LS. Nichols LA, Norris SH, Rothlein R,
Scharschmidt LA, Lipskc PE: Treatment of refractory rheumatoid
arthritis with a monoclonal antibody to intercellular adhesion
molecule 1. Arthritis Rheum 37:992-999. 1994
Документ
Категория
Без категории
Просмотров
3
Размер файла
958 Кб
Теги
necrosis, factors, endothelial, vascular, monoclonal, arthritis, antibody, rheumatoid, antitumor, deactivation
1/--страниц
Пожаловаться на содержимое документа