close

Вход

Забыли?

вход по аккаунту

?

Use of a radiolabeled monoclonal antibody against e-selectin for imaging of endothelial activation in rheumatoid arthritis.

код для вставкиСкачать
ARTHRITIS & RIIEUMA.I'ISM
Vol. 39, No. 8, August 1996, pp 1371-1375
Q 1996, American College of Rheumatology
1371
USE OF A RADIOLABELED MONOCLONAL ANTIBODY AGAINST
E-SELECTIN FOR IMAGING OF ENDOTHELIAL ACTIVATION I N
RHEUMATOID ARTHRITIS
PETER T. CHAPMAN, FRANCOIS JAMAR, EDWARD T. M. KEELAN,
A. MICHAEL PETERS, and DORIAN 0.HASKARD
Objective. To determine the potential of "'Inlabeled anti-E-selectin monoclonal antibody (MAb) to
image localized endothelial activation in rheumatoid
arthritis (RA).
Methods. Fourteen patients with RA were studied
after intravenous administration of "'In-labeled
F(ab'), fragments of MAb against the cytokheinducible endothelial cell activation antigen E-selectin
(MAb 1.2B6). To compare uptake of 1.2B6 with that of
nonspecific immunoglobulin, "'In-labeled polyclonal
human immunoglobulin (HIG) was separately administered to 6 of these patients and the relative uptake of
each tracer was determined.
Results. Prominent and discrete uptake of the
radiolabeled MAb 1.2B6 was clearly visible in inflamed
joints of all patients. Compared with "'In-HIG, "'In1.2B6 provided superior images in terms of sensitivity
and image intensity. Furthermore, the distribution of
uptake in inflamed joints was different for the 2 tracers,
with 1.2B6 showing a more focal localization in synovium.
Conclusion. This study demonstrates that it is
possible to objectively assess E-selectin expression on
activated endothelium in vivo in patients with RA, using
Supported by a grant from Hammcrsmith Hospitals NHS
Trust. Dr. Chapman's work was supported by the Dorothy Eden
Fellowhip, Arthritis Foundation of New Zealandikthritis and Rheumatism Council and Medical Research Council (UK). Dr. Jamar's
work was supported by the Foundation St-Luc, and the European
Economic Community, Brussels, Belgium.
Peter T. Chapman, BSc, FRACP, FranGois Jamar, MD,
Edward T. M. Keclan, MKCP, A. Michael Peters, MD, MRCP. Dorian
0. Haskard, DM, FRCP: Koyal Postgraduate Medical School, Hamrnersmith Hospital, London, UK. Drs. Chapman and Jamar contributed equally to this study.
Address reprint requests to Dorian 0. liaskard, DM, FRCP,
Rhcumatolog Unit, Department of Medicine, Royal Postgraduate
Medical School, I Iammersmith Hospital. Du Canc Road, 1.ondon
W12 ONN, UK.
Submitted for publication June 16, 1995; accepted in revised
form March 5, 1996.
a radiolabeled MAb. This technique has considerable
potential for monitoring disease activity and response to
therapy in inflammatory diseases.
Improved understanding of basic mechanisms of
inflammation over the last decade has led to an increased awareness of the central and active role of
the vascular endothelium in the control of the inflammatory response. Following exposure to cytokines such
as interleukin-l and tumor necrosis factor a,endothelial
cells (EC) become activated and exhibit a complex
proinflammatory response which includcs the synthcsis
and secretion of cytokines and growth factors, together
with an enhanced adhesiveness for circulating leukocytes. The enhanced adhesiveness of cytokine-activated
EC for leukocytes is largely related to the induced
surface expression of adhesion molecules such as Esclectin, intercellular adhesion molecule l, and vascular
ccll adhesion molecule 1 (1).
The recognition of the importance of endothelium in inflammation has fostered a nccd for clinical
techniques to study local and systemic endothelial activation. E-selectin is a single-chain glycoprotein which is
expressed by activatcd, but not unstimulated, E C and is
not thought to be present on other cell types. Its
expression on the luminal surface of vascular endothelium contributes to the rolling of leukocytes prior to
their emigration into tissues. Leukocytes which bind
E-selectin include ncutrophils, eosinophils, basophils,
monocytes, and a subpopulation of memory T cells.
Immunohistologic studies have revealed E-selectin expression in a wide variety of acute and chronic inflammatory conditions (2). In view of these propertics,
E-sclectin presents a promising target for the in vivo
detection of endothelial activation.
Accordingly, we havc demonstrated, in preclinical studies on pigs, that radiolabeled anti-E-selectin
CHAPMAN ET AL
1372
monoclonal antibody (MAb) 1.2B6 specifically binds
activated endothelium in vivo and can be used to image
arthritis (3,4). Moreover, in view of potential clinical
advantages compared with the use of the whole MAb
molcculc, wc have more recently cvaluatcd radiolabcled
F(ab')2 fragments of 1.2B6 and have obtained similarly
encouraging results (5,6). We now report our first clinical cxpcrience using radiolabeled F(ab'), fragments of
1.2B6 in patients with RA, and comparing uptake of
1.2B6 with that of radiolabeled polyclonal human immunoglobulin (HIG), the use of which is an established
technique for the evaluation of disease activity in RA (7).
PATIENTS AND METHODS
Patients. The study was approved by the Royal Postgraduate Medical School and Hammersmith and Queen Charlotte's Special Health Authority Research Ethics Committee.
All patients gave written informed consent. Fourteen patients
who fulfilled the American College of Rheumatology (formerly, the Amcrican Rheumatism Association) 1987 criteria for
rheumatoid arthritis (RA) (8) were studied. Clinical disease
activity was assessed on the first day of the study by scoring 20
joints on a scale of 0-3 (0 = none; 1 = mild; 2 = moderate; 3
= marked) for tenderness (Kitchie articular index) (9) and
swelling (10). Routine laboratory parameters that were recorded included hematology profile, erythrocyte sedimentation
rate (ESR), C-reactive protein (CRP) level, and rheumatoid
factor status. Circulating E-selcctin in serum obtained just
prior to MAb administration was measured by enzyme-linked
immunosorbent assay (ELISA) as previously described (11).
The human anti-mouse antibody (HAMA) response was determined by E I S A (12), comparing blood samples obtained
pre- and 12-14 days post-injection of the MAb.
Monoclonal antibody preparation and radiolabeling.
The anti-E-selectin MAb 1.2B6 is a mouse IgGl generated
against cytokine-activated human EC (13). The F(ab'), fragment was obtained by pepsin digestion, after which immunoreactivity was confirmed by ELISA on cultured EC. Radiolabeling of the F(ab'), fragment (35 pg) and of polyclonal
human IgG (Sandoglobulin, 1 mg; Sandoz, Camberley, UK),
both with 20-25 MBq "'In, was performed on the day of each
experiment, as previously reported ( 5 ) . The radiopharmaceutical purity of radiolabeled immunoglobulin preparations was
>98% in all cases, as assessed by instant thin-layer chromatography. 'I'he mean i- SD injected dose was 13.4 2 0.8 MBq.
Prior to use in this study, the MAb 1.2B6 preparation was
subjected to extensive safety and toxicity testing at independent laboratories.
Imaging. Each of the 14 patients underwent intravenous injection with "'In-labeled 1.2B6. Six of them (paticnts
9-14), all with stable RA as judgcd clinically and by serial ESR
and CRP measurements performed at the time of each scan,
reccivcd '"In-1.2B6 and "In-HIG administered separately, 2
wccks apart in random order.
Images o f the peripheral joints and anterior and posterior views of the thorax and abdomen were obtained 4 hours
and 24 hours post-injection of MAb, using a large-field-of-
'
view gamma camera equipped with a medium-energy generalpurpose collimator. Scintigraphie joint uptake was evaluated
without knowledge of the clinical scores, using a 4-point visual
scoring system (0 = no uptake; 1 = minor uptake; 2 =
moderate uptake; 3 = marked uptake). The clinical and
scintigraphic joint scores were determined on the following
joints: shoulders, elbows, wrists, metacarpophakmgeal joints
(as a group), proximal and distal interphalangeal joints (as a
group), knees, ankles, midtarsal joints, and mctatarsophalangeal joints (as a group). Thus, a total of 20 joint scores were
recorded for each patient, allowing a maximum global score of
60. For comparative purposes, scintigraphic uptake of "'In1.2B6 or "'In-HIG was also measured as pixel counts in the
region of interest (joint space) of individual joints, and expressed as a jointsoft tissue ratio (joint space pixel counts:
adjacent soft tissue pixel counts) after correction for background activity. Only joints that were judged to be positive with
both tracers were included in this latter analysis.
Statistical analysis. Unless stated otherwise, data are
presented as the mean 2 SD. Correlations between clinical
scores, scintigraphic scores, and biochemical data wcre established using linear regression analysis. Discordances between
scintigraphic scores (1.2B6 versus HIG) or between scintigraphic and clinical scores were analyzed using the McNemar
test for paired data. The semiquantitative image contrast data
for 1.2B6 and HIG were compared using Student's paired t
test. P values less than 0.05 were considered significant.
RESULTS
Prominent and discrctc accumulation of the radiolabeled anti-E-selectin MAb 1.2B6 was easily detectable in inflamed joints of all 14 RA patients. Joint
uptake was clearly visible at 4 hours, and there was an
increase in image intensity in the same joints by 24
hours. Image resolution was sufficient to distinguish
uptake by individual joints, including thc small joints of
the hands and feet (Figures 1A and C). In the cohort of
patients, global scintigraphic scores correlated significantly with both global joint tenderness scores (r = 0.83,
P < 0.001) and global joint swelling scores (r = 0.65, P <
0.01). However, the correlations wcrc not absolute, and
positive imagcs wcrc somctimcs obtained in joints that
clinically appeared to be inactive. Importantly, scvcral
joints with marked degenerative changes without clinical
inflammation showed no antibody uptake, suggesting
that the technique was not detecting secondary osteoarthritis.
Among thc 6 patients (paticnts 9-14) who separately received "'In-HIG in addition to anti-E-sclectin
1.2B6 in ordcr to compare thc relative capacity of thc 2
agents to localize in inflamed joints, radiolabeled HIG
was takcn up in most inflamed joints, particularly in the
presence of soft tissue swelling. Although there was a
degree of concordance between the 2 tracers (80% and
IMAGING ENDOTHELIAL ACTIVATION IN RA
Figure 1. Gamma camera images of the hands and feet of a paticnt
with longstanding rhcumatoid arthritis, taken 24 hours aftci intravenous injection of separately administered "'In-labclcd anti-E-sclcctin
monoclonal antibody 1.2B6 (A and C) and "'In-labeled polyclonal
human immunoglobulin (B and D). Although images with both tracers
demonstrate a similar distribution of uptake, uptake of 1.2B6 is clearly
morc intense, with a more focal pattern and a higher targct-tobackground ratio.
76%) concordance of individual joint scores at 4 hours
and 24 hours, respectively), positive images were obtained with 1.2B6 in a number of joints that were
negative for HIG (at 4 hours, positive 1.2B6 images and
negative HIG imagcs in 19% of the joints studied with
both tracers; at 24 hours, positive 1.2B6 imagcs and
negative HIG images in 22%). Conversely, positive HIG
and negativc anti-E-selectin images occurred in only 1%
(4 hours) and 2% (24 hours) of all joints studied. Joints
that localized 1.2B6 but not HIG tended to bc characterized by an absence of clinically detectable joint swelling. When the data wcrc vicwcd as a whole, the 1.2B6
scan gave equivalent joint scores at 4 hours and higher
joint scores at 24 hours, compared with clinical examination ( P < 0.005 for both joint swelling and joint
tenderness, by McNemar tcst), suggesting that the traccr
was identifying subclinical synovitis. In comparison, the
HIG scan gave lower scores than thc clinical indices at 4
hours ( P < 0.05 and 1' < 0.001 for joint swelling and
joint tcndcmcss, respectively, by McNemar test), and
equivalent scores at 24 hours.
Apart from bcing more sensitive than "'Inlabeled HIG in detecting inflamcd joints, lllTn-labeled
1373
1.2B6 yielded images that were more intense and more
focal than those obtained with "'In-HIG, in almost all
joints studicd (Figurc 1). In thc joints in which positivc
imagcs wcrc obtained with both tracers, jointsoft tissue
ratios at 4 hours and 24 hours were 2.0 -+ 0.9 and 2.8 %
1. l , respcctivcly, for "'In-HIG, and 2.8 t 1.4 and 5.0 t
2.8, respectively, for '"In-1.2B6 (I' < 0.005 at 4 hours,
P < 0.0001 at 24 hours, by Student's paired t test).
Thc superior synovial localization of anti-Eselectin MAb 1.2B6 is illustratcd in Figure 2, in which
gamma camera images of a patient with bilateral effusions due to chronic synovitis arc shown. Prominent
focal uptake of ''lln-1.2B6 is seen particularly in the
rcgion of the synovium of both knees, with a more
diffuse and less intense pattern of uptake also evident in
thc surrounding articular area (Figure 2A). In comparison, the "'In-HIG image demonstrates a diffuse articular uptake but without prominent focal synovial uptakc
(Figurc 2B). In the same patient, synovial fluid (SF)
sampling on the day of each scan showed that the 2
tracers accumulated in the SF in equivalent quantities
(0.02% injected dose/ml SF) and with similar SF:plasma
ratios (HIG 0.49, 1.2B6 0.53). Thus, the increased
uptake of E-sclcctin compared with HIG as seen by
scintigraphy must rclatc to spccific uptakc by synovial
tissue, as opposed to nonspecific exudation into SF.
A summary of paticnt demographic data, global
clinical and scintigraphic scores, and clinical indices of
the acute-phase rcsponsc (CRP, ESR) is given in Table
1. Global scintigraphic scores correlated with the ESR
for both tracers ( P <0.05), but not with the CRP. Thcrc
was no correlation between the global clinical scores and
laboratory indices of inflammation, or bctwcen scintigraphic anti-E-selcctin MAb joint uptake and circulating lcvels of soluble E-selectin. No HAMA wcrc dctcctable at 12-14 days following MAb injection.
DISCUSSION
This is the first publishcd dcscription of the
clinical usc of a radiolabeled MAb against E-selcctin to
localize endothelial activation in inflamed tissues. Wc
found that the antibody gavc distinct images of inflamed
joints in patients with rhcumatoid arthritis. Moreover,
positive images of "clinically silent" joints, suggesting the
presencc of E-selectin in the undcrlying synovium, were
frequently obtained. Although MAb 1.2B6 uptake was
usually visiblc in inflamed tissues within 4 hours of
injection, images were clearly improved at 24 hours, and
also when obtained at 48 hours. We attribute this in part
to the contrast created by clearance of background
1374
CHAPMAN ET AL
Figure 2. Gamma camera images of t h e knees of a patient with rheumatoid arthritis, taken 24 hours after
intravenous injection of separately administered '"In-labeled anti-E-selectin monoclonal antibody 1.2B6 (A) and
l'lln-labeled polyclonal human immunoglobulin (HIG) (B).Both knees, particularly the right knee, had clinically
active disease with synovial effusions. Uptake of 1.2B6 is most prominent in the region of the synovium (arrow),
although a less intense and more diffuse uptake in the surrounding articular area is also seen. In contrast, the HIG
uptake shows a diffuse distribution without prominent synovial localization.
radioactivity ( 5 ) ,and in part to the progressive internalization of antibody-antigen complexes into endothelial
cells that has been shown to occur in vitro (14).
Having determined its feasibility for imaging
rheumatoid joints, we next compared radiolabeled antiE-selectin MAb 1.2B6 with radiolabelcd HIG, an established agent for imaging of RA (7). Although the precise
mechanism of uptake of HIG is uncertain, it is generally
believed that this tracer localizes by nonspecific diffusion
across compromised endothelium. We found that antiE-selectin provided superior images of inflamed joints in
terms of sensitivity and image intensity. Furthermore,
the distribution of uptake in inflamed joints was different
for the 2 tracers, with 1.2B6 showing a more focal
localization in synovium. Moreover, the ability of antiE-selectin to localize in joints that had no clinically
Table 1. Characteristics of the rheumatoid arthritis patients studied*
~
Paticnt
AgeISex
Global joint
tenderness score,
0-60
1
2
3
4
5
6
7
8
Y
10
11
12
13
14
64/F
45F
h3lM
651F
64iF
55iF
57iF
601F
71/F
77lF
681M
59/F
651F
63F
12
3
17
14
4
30
16
26
12 (14)
(8)5
8 (11)
(4) 6
10 (12)
(17) 15
Global joint
swelling score,
0-60
19
5
21
19
Y
23
12
15
11 (12)
(10) 10
6 (9)
(8) 10
8 (10)
(15) 14
ESR,
mmlhour
CRP,
mgilitcr
Global
scintigraphic score,
0-60
Circulating
sE-selectin,
ndmlf
30
24
18
46
20
55
26
50
55 (43)
(40) 40
31 (42)
(44) 30
62 (78)
(80) 74
c2
12
9
17
5
66
29
67
36 (29)
(177) 107
c 2 (33)
4
46 (38)
(72) 8Y
25
7
29
27
13
44
25
34
17 (11)
(13) 23
27 (10)
( 8 ) 16
31 (19)
(21) 34
75
36
41
58
38
95
111
39
109
91
22
60
I47
153
* All 14 patients underwent E-selectin scans. Paticnts 9-14 also underwent polyclonal human immunoglobulin (HIG) scans; the 2 scans wcrc
performed in random ordcr 2 weeks apart. Values in parenthescs are those obtained on the day of HIG scanning; all other values are thosc obtained
on the day of E-selectin scanning. ESR = erythrocyte sedimentation rate; CRP = C-reactive protein.
t The mean 2 SD level of circulating soluble E-selectin (sE-selectin) in 160 pooled normal control sera was 83 ? 4.3 ngiml.
IMAGING ENDOTHELIAL ACTIVATION IN RA
detectable soft tissue swelling or HIG uptake strongly
suggests that the uptake is not dependent on nonspecific
protein exudation. Although in some joints there is a
nonspecific component to 1.2B6 accumulation (as shown
by its presence in SF), the data are consistent with the
notion that enhanced MAb 1.2B6 uptake is predominantly due to binding to E-selectin that has been induced
on activated vascular endothelial cells at sites of inflammation, as has been previously demonstrated in preclinical studies (3).
Although there are a number of clinical imaging
techniques for localizing inflammation (15), the use of
anti-E-selectin MAb has the advantage of targeting a
well-characterized endothelial activation antigen. The
presence of E C activation in almost all forms of acute
and chronic inflammation highlights the versatility and
potential of this technique in the clinical setting. While
the place of anti-E-selectin imaging in the clinical
evaluation and monitoring of patients with RA remains
to be determined in the context of other newly developed imaging modalities (16), we believe it will provide
useful objective criteria for the initial assessment of
patients and for the monitoring of outcomes of established and novel therapies.
In summary, we have shown that a radiolabeled
antibody against E-selectin can be used to image localized synovitis in patients with RA, and that it is clearly
superior to radiolabeled nonspecific immunoglobulin in
terms of sensitivity, image intensity, and focal localization in synovium. This technique has considerable promise for the evaluation of endothelial activation in patients with inflammatory arthritis and other forms of
inflammatory disease.
3.
4.
5.
6.
7
8
9
10
11
12
ACKNOWLEDGMENTS
We are indebted to Prof. J. P. Lavender for his
foresight in initiating discussions which wcrc a catalyst for this
work. We are very grateful to the radiographers in the Nuclear
Medicinc Unit, Hammersmith Hospital; and for the helpful
advice of Dr. Neil Turner (preparation of MAb 1.2B6 F[ab’],
fragments), Dr. Justin Mason (circulating soluble E-sclectin
ELISA), and Dr. Robert Spooner (HAMA assay).
13
14.
REFERENCES
15.
1. Bcvilacqua MP: Endothelial-leukocyte adhesion molcculcs. Annu
Rev Immunol 11:767-804, 1993
2. Mason JC, Haskard DO: The clinical importance o f lcucocytc and
16.
1375
endothelial cell adhesion molecules in inflammation. Vascular
Mcd Rev 5:249-275, 1994
Kcclan ETM, Licence ST,Peters AM. Binns KM, Haskard DO:
Characterization of E-selectin expression in vivo with use of a
radiolabcllcd monoclonal antibody. Am J Physiol 266:H279H290, 1994
Keelan ETM, Harrison AA. Chapman PT, Binns RM, Peters AM,
IIaskard DO: Imaging vascular endothelial activation: an approach using radiolabeled monoclonal antibody against the cndothelial cell adhesion molecule E-selectin. J Nucl Med 35:276-281,
1994
Jamar F, Chapman PT, Harrison AA, Hinns RM, Ilaskard DO,
Peters AM: Imaging endothelial cell activation in inflammatory
arthritis using an In-1 11 labeled anti-G-sclectin monoclonal
F(ab’)2. Radiology 1945343-850, 1995
Chapman PT, Jamar F, Harrison AA, Binns RM. Peters AM,
Ilaskard DO: Noninvasive imaging of E-selectin expression by
activated endothelium in urate crystal-induced arthritis. Arthritis
Rheum 37:1752-1756, 1994
Breedveld FC, van Kroonenburgh MJPG, Camps J N , Feitsma
IIIJ. Markussc HM, Pauwcls BKJ: Imaging of inflammatory
arthritis with technetium-9Ym labelled IgG. J Nucl Med 30:20172021, 1989
Arnett FC, Edworthy SM, Hloch DA, McShane DJ, Fries JF,
Cooper NS, Hcalcy I.A, Kaplan SK, 1,iang MH, 1,uthra €IS,
Medsgcr TA Jr, Mitchell DM, Ncustadt DH, Pinals KS, Schallcr
JG, Sharp JT, Wilder RL, IIunder GG: The American Rheumatism Association 1987 rcviscd criteria for the classification of
rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988
Ritchic DM, Hoylc JA, McInncs JM, Jasani MK, Ilalakos TG,
Grievcson P. Buchanan WW: Clinical studies with an articular
index for the assessment of joint tenderness in patients with
rheumatoid arthritis. QJM 37:393-406, 1968
Williams HJ, Ward JK, Reading JC, Hgger MJ, Grandone JT,
Samuelson CO, Furst DG, Sullivan JM, Watson MA, Guttadauria
M, Cdthcart ES, Kaplan SB, IIalla JT, Weinstein A, Plotz PH:
Low-dose D-penicillamine therapy in rheumatoid arthritis: a controlled, double-blind clinical trial. Arthritis Rheum 26581-592,
1983
Montefort S, Lai CK, Kapahi P, L u n g J, Lai KN, Chan HS,
Haskard 110, IIowdrth PII, Ilolgate ST: Circulating adhesion
molcculcs in asthma. Am J Kcspir Crit Care Med 1491149-1152,
1994
Courtcnay-Ixk NS, Epenetos AA, Moore R, Larchc M, Pcctasides D, Dhokia €3, Kittcr MA: Development of primary and
secondary immune responses to mouse monoclonal antibodies
used in the diagnosis and therapy of malignant ncoplasms. Cancer
Rcs 46:6489-6493, 1986
Wcllicome SM, ’Ihornhill MH, Pitzalis C, Thomas S, Lanchbury
JS, Panayi GS, IIaskard DO: A monoclonal antibody that detects
a novel antigen on endothelial cells that is induced by tumor
necrosis factor. IL-1 or lipopolysaccharidc. J Immunol 144:25582565, 1990
Von Asmuth H U , Smccts EF, Ginsel [*A, Ondenvater JJM,
Ixeuwcnbcrg JFM, Huurman W A Evidence for cndocytosis of
E-sclcctin in human endothelial cells. Eur J Immunol 22:25192.526, 1992
Gardncr 1’. Ostcr ZH: Rubor, dolor, calor, tumor and radionuclide
scans. N Engl J Med 321:970-972, 1989
De Bois MHW. Pauwels EKJ, Brccdvcld FC: New agents for
scintigraphy in rheumatoid arthritis. Bur J Nucl Mcd 22:13391346, 19%
Документ
Категория
Без категории
Просмотров
2
Размер файла
529 Кб
Теги
endothelial, radiolabeled, monoclonal, imagine, activation, arthritis, antibody, use, rheumatoid, selecting
1/--страниц
Пожаловаться на содержимое документа