|
|
||||||||
Published online before print March 12, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
4 integrin subunit attenuates the acute inflammatory response to stent implantation yet is insufficient to prevent late intimal formation
Vascular Biology Laboratory, Division of Cardiology, University of Ottawa Heart Institute, Ontario, Canada
1 Correspondence: Vascular Biology Laboratory, Division of Cardiology, University of Ottawa Heart Institute, Ottawa, Ontario, Canada K1Y4W7. E-mail: eobrien{at}ottawaheart.ca
| ABSTRACT |
|---|
|
|
|---|
4 subunit of very late antigen-4 is sufficient for the late attenuation of stent intimal area (IA). Twenty-four hypercholesterolemic rabbits underwent iliac artery balloon injury, followed 2 weeks later by stent implantation, and the animals were randomized to receive an anti-
4 antibody (HP1/2) or a nonspecific isotypic control immunoglobulin (1E6) intravenously 1 h before stenting. Compared with controls, HP1/2-treated rabbits showed 50%, 51%, and 44% reductions in the percentage on intimal cells that were macrophages on days 3, 7, and 28 after stenting and a 59% reduction in intimal proliferation on day 3. Although stent IA was reduced by 63% and 48% in the antibody-treated group compared with the control group on days 3 and 7, this difference was not present on day 28. These data highlight the need for sustained, anti-inflammatory therapies for the prevention of stent intimal formation.
Key Words: restenosis leukocyte adhesion
| INTRODUCTION |
|---|
|
|
|---|
4ß1, is a member of the ß1-integrin subfamily and is expressed on resting monocytes, lymphocytes, neutrophils, basophils, and eosinophils [7
]. In vivo studies using blocking monoclonal antibodies (mAb) and inhibitory peptides demonstrate that VLA-4 plays a critical role in the early stages of vascular inflammation [7
, 8
]. VLA-4 mediates cell adhesion via its counter-receptor vascular cell adhesion molecule-1 (VCAM-1) found on endothelial cells and activated smooth muscle cells (SMCs) [9
, 10
].
HP1/2 is a mAb that binds to the
4 subunit of VLA-4 [7
]. Studies of vascular lesion formation in mice, rabbits, pigs, and baboons document that blocking VLA-4 is highly efficacious in reducing vessel wall inflammation [11
12
13
14
15
]. Given the role of inflammation in the response to stent implantation, we sought to determine in a rabbit model if acute antagonism of the
4 integrin subunit with the HP1/2 mAb would not only attenuate inflammation but also be sufficient to prevent stent intimal formation 28 days later.
| MATERIALS AND METHODS |
|---|
|
|
|---|
4 subunit of the human VLA-4 and cross-reacts with the respective rabbit homologue on mononuclear leukocytes but does not bind to rabbit neutrophils [13
, 16
, 17
]. The mAb 1E6 (Biogen Inc.), a mouse anti-human lymphocyte function-associated antigen-3 IgG1, was used as an isotypic (negative) control antibody. Euthanasia was performed on days 3, 7, and 28 poststenting. Intimal inflammatory cell abundance as well as proliferation and stent intimal area (IA) were tabulated.
Vascular injury and stent implantation
All animal procedures were performed with the approval of the University of Ottawa Animal Care Committee (Ontario, Canada) and followed the guidelines of the Canadian Council on Animal Care. Twenty-four male New Zealand white rabbits (3.03.5 kg, Charles River Laboratories, Quebec, Canada) were studied. Beginning 1 week before balloon injury, rabbits received a 0.3% cholesterol diet (Harlan Teklad, Madison, WI). Under general anesthesia with ketamine [25 mg/kg, intramuscularly (i.m.)], medazolam (24 mg/kg, i.m.), and isoflurane (via an endotracheal tube), balloon injury was conducted on both iliac arteries via the carotid artery [18
]. After waiting 2 weeks for re-endothelialization to occur, a stainless-steel stent (S670 stent, 3.0 mm diameterx24 mm length, Medtronic AVE, Inc., Santa Rosa, CA) was deployed at six atmospheres in each balloon-injured iliac artery via the noninstrumented, contralateral carotid artery [19
, 20
]. In separate experiments, we observed VCAM-1 expression in re-endothelized rabbit iliac arteries 2 weeks postballoon injury (data not shown). Rabbits were blindly randomized to receive a 3-mg/kg single intravenous (i.v.) bolus of the HP1/2 or 1E6 mAb 1 h before stent implantation. All animals received Heparin (125 U/kg, Leo Pharma Inc., Ajax, Ontario, Canada) as an i.v. bolus at the outset of the procedure. To limit stent thrombosis, all rabbits were given acetylsalicylic acid (rectal gel, 10 mg/kg, per os) every day, starting 3 days before balloon injury and continuing until euthanization. As per usual clinical practice, the complementary antiplatelet agent clopidogrel bisulfate (Sanofi-Synthelabo Canada, Montreal, Quebec) was administered transdermally, beginning with a loading dose of 4 mg/kg on the day before stenting and continued as 1 mg/kg/day thereafter until euthanasia.
Tissue harvest and analysis
Three, 7, or 28 days after stent implantation, animals were killed using i.v. euthanyl (Schering-Plough, Quebec, Canada). All animals received bromodeoxyuridine (BrdU; 50 mg/kg, i.v., Sigma Chemical Co., St. Louis, MO) 1 h before euthanasia to allow immunolabeling of proliferating cells. After exposing the aorta-iliac system, normal saline was used to flush off residual blood. The stented arteries were serially cut into five 4.8 mm-long rings, which were processed as follows:
One ring was embedded in methylmethacrylate after overnight fixation with 10% neutral-buffered formalin. Cross-sections (5 µm-thick) were cut with a D-Profile tungsten carbide knife (Delaware Diamond Knives Inc., Wilmington), and hematoxylin/eosin as well as Movat pentachrome-stained slides were obtained.
One ring was stored in liquid nitrogen immediately after dissection and reserved for future studies.
One ring was opened and embedded in optimal cutting temperature (Miles Inc., Elkhart, IN) after the stent was carefully removed by manual dissection. Frozen cross-sections (5 µm) were cut and were stained with hematoxylin/eosin or processed for immunohistochemistry.
One ring was opened, and the stent was manually removed before serial 5 µm cross-sections were cut at subsegment intervals of 50 µm. For morphometric analyses, nine cross-sections from three subsegments (i.e., three cross-sections per subsegment) were examined. The lumen area (LA) as well as the area circumscribed by the internal elastic lamina (IEL) were measured on Movat pentachrome-stained sections using a computer-assisted digital system (Image-Pro Plus, Media Cybernetics, Silver Spring, MD). IA was defined as: IA = IEL area LA. For each specimen, the extent of arterial wall injury was graded using a standardized protocol developed by Schwartz and colleagues [21 ], which tabulates in a semiquantitative manner the degree of injury imparted to the vessel wall at each stent strut. For example, a score of zero corresponds with an-intact, internal elastic, and the maximum score (three) is defined by medial laceration that extends through the external elastic lamina. Similarly, the extent and density of inflammatory cells surrounding each stent strut were scored using a system devised by Kornowski et al. [5 ]. For example, a score of zero is assigned if no inflammatory cells are present, and the maximum score of three is used if there is a dense, circumferential collection of lymphohistiocytic cells around a strut.
A stent segment devoted to scanning electron microscopy (SEM) was cut open longitudinally, partially flattened, and fixed in 1.6% glutaraldehyde before being dehydrated and dried with liquid CO2. The samples were coated with gold and examined using SEM (XL 30 ESEM, Philips Electronics Ltd., Markham, Ontario, Canada). SEM photomicrographs of each specimen were specifically examined for adherent leukocytes as well as reconstitution of the endothelium.
Immunohistochemistry
Immunohistochemistry was performed on paraffin-embedded and fresh-frozen stent intimal tissue specimens. The following primary antibodies were used: RAM-11 for macrophages (titer 1:50, Dako, Mississauga, Ontario, Canada), anti-BrdU (titer 1:50, Dako) for proliferating cells, anti-SMC
-actin (titer 1:400, Sigma Chemical Co.) for smooth muscle cells/myofibroblasts, RPN3/57 (titer 1:25, Serotec, Raleigh, NC) for neutrophils, and anti-CD43 (titer 1:50, Serotec) for T cells. After incubating with the primary antibody, a species-specific, biotinylated secondary antibody was applied, followed by incubation with an avidin-biotin peroxidase complex (VECTASTAIN Elite ABC kit, Vector Laboratories, Burlingame, CA) and visualization with 3,3-diaminobenzidine (Sigma, Oakville, Ontario, Canada). Slides were counterstained with hematoxylin. Immunolabeled macrophages, T cells, neutrophils, and proliferating cells in the intima were counted and expressed as a percentage of the total number of intimal cells per high-power field (HPF; magnification, x400). For each artery, nine cross-sections were examined, using three HPFs per cross-section.
Blood analyses
Blood samples were drawn before and 1 h after each antibody was administered as well as at euthanasia for lipid profile measurements and serum HP1/2 levels. A cell-free electrochemiluminescence competition assay using ruthenium-labeled HP1/2 was used to determine serum levels of HP1/2 [22
].
Statistics
All data are presented as mean ± SEM. A one-way ANOVA was used for multiple comparisons between groups, and a paired t-test was used for comparisons between two treatment groups. Nonsignificance was defined by a P value <0.05.
| RESULTS |
|---|
|
|
|---|
Inflammatory response to stenting
Rabbits receiving the control Ig (1E6) showed progressive, intimal accumulation of RAM-11-positive monocytes/macrophages. Compared with controls, rabbits receiving the HP1/2 treatment showed 50%, 51%, and 44% reductions in the number of monocytes/macrophages present in the stent intima on days 3, 7, and 28, respectively (e.g., percentage monocyte/macrophages per intimal HPF: 7.35±1.47% vs. 3.68±0.86%, 16.88±3.67% vs. 8.30±2.04%, 39.75±6.82% vs. 22.19±4.36%, respectively; P<0.05 for all intervals; Fig. 1A
and 1B
). SEM also demonstrated fewer mononuclear leukocytes adherent to the surface of stented iliac arteries in the HP1/2-treated group compared with controls (Fig. 2
). Moreover, compared with the control rabbits, the endothelium of the HP1/2-treated rabbits showed more complete restitution in the stented segments and fewer gaps between endothelial cells. On day 3, there was a 37% reduction in the inflammatory score, an index of the abundance of inflammatory cells immediately adjacent to the stent struts (e.g., 1.39±0.22 vs. 2.22±0.17, P=0.009). However, on days 7 and 28 after stenting, the inflammatory score was similar for the HP1/2 and 1E6 groups (e.g., 1.61±0.32 vs. 1.76±0.25, P=0.72; 1.86±0.30 vs. 1.69±0.21, P=0.65, respectively). The abundance of intimal T cells and neutrophils was similar in both groups (see http://www.jleukbio.org for supplemental figure).
|
|
|
| DISCUSSION |
|---|
|
|
|---|
4 integrin at time of stent implantation is sufficient not only to reduce the acute inflammatory response to stent implantation but also to attenuate the late intimal response. Our results indicate that a single i.v. dose of the antibody HP1/2 to the
4 integrin subunit results in therapeutic antibody levels for at least the first 3 days postadministration and acutely attenuates inflammatory cell infiltration and proliferation in the stented vessel wall. However, by day 28, stent IA in the HP1/2-treated rabbits was indistinguishable from that of the control rabbits.
Our current understanding of how ISR lesions form is derived from scattered clinical and experimental observations [23
]. Within hours to days after a stent is implanted, platelets with organizing thrombi may serve as a network for the attachment of fibrin and mononuclear cells that initiate an inflammatory response, particularly in the tissue adjacent to the stent struts. This early inflammatory response may elicit the release of a series of growth and chemotactic factors crucial for the recruitment of additional mesenchymal cells that later differentiate into
-actin-positive SMCs. The origin of the mesenchymal cells that become organized within ISR lesions is unclear. The hypothesis most commonly accepted is that ISR tissue is a result of excessive growth and inward migration of vascular SMCs or myofibroblasts from the media and/or adventitia. In other words, it is presumed that the mesenchymal cells that populate ISR lesions are arterial wall myofibroblasts or SMCs. However, the possibility that pluripotent blood-borne cells and proteins accumulate along the inside of the stent struts and eventually block the stent cannot be ruled out [24
, 25
]. Regardless, the key, initiating event in the genesis of stent intima formation involves inflammatory cells.
In the current study, rabbits treated with the anti-
4 mAb had 50%, 51%, and 44% fewer mononuclear cells in the stent intima on days 3, 7, and 28, respectively. These results are consistent with rabbit and baboon arterial injury experiments not involving stent implantation, where administration of the same anti-
4 antibody reduced the intimal influx of monocytes by 70% and 53%, respectively [13
, 15
]. Moreover, our results share similarities with those obtained when blockade of Mac-1-dependent leukocyte adhesion to fibrinogen and platelets is used to prevent stent intimal formation. Fourteen days after stent insertion in normocholesterolemic rabbit arteries, Rogers and colleagues [26
] observed a 38% reduction in IA with repetitive injections of an anti-Mac-1 antibody, despite no effect on intimal proliferation. It should be noted, however, that with Mac-1 inhibition, there is the theoretical concern of impairing host defenses and creating a predisposition to infection, as adhesion of not only monocytes but also neutrophils occurs.
Although there were still a significant number of inflammatory cells involved in the intimal response to stent implantation in the HP1/2-treated rabbits in our study, the reduction in stent IA is particularly impressive when one considers that the serum cholesterol levels in these rabbits were
15x baseline levels. Independent of the inflammatory response to balloon injury and stenting, marked hyperlipidemia alone is proinflammatory and dramatically up-regulates the expression of VCAM-1 [27
28
29
30
]. The attenuation of intimal inflammatory cell involvement seen in this study translated into a 37% reduction in the abundance of inflammatory cells immediately adjacent to the stent struts 3 days poststenting. However, at later intervals, this peristrut reduction of inflammation was not observed, most likely as a result of the limited therapeutic window of a single injection of the HP1/2 mAb and the sustained inflammatory reaction provoked by the stainless-steel stent. The concentration of HP1/2 mAb returned to baseline levels sometime between days 3 and 7 poststenting, and despite 64% and 43% reductions in stent IA on days 3 and 7 in the HP1/2-treated versus control rabbits, stent IA on day 28 was similar in both groups. As the intention of the study was to determine if acute antagonism of VLA-4 at time of stent implantation is sufficient for late prevention of stent intimal formation, a strategy that might be easily translated into clinical practice, repetitive injections of HP1/2 were not pursued. The possibility that HP1/2 might have had effects beyond inhibiting mononuclear cell involvement in lesion formation cannot be excluded in this experiment. For example, observations in vitro as well as in vivo suggest that the interaction between VLA-4 and VCAM-1 is a prerequisite for phenotypic modulation of SMCs during physiological blood vessel development as well as atherogenesis [31
]. Moreover, as Rabinovitch and colleagues [8
] demonstrated, antagonism of VLA-4 with a synthetic-connecting segment-1 peptide of fibronectin inhibits intimal formation in rabbit cardiac allografts. Hence, in this study, there may also be a direct, anti-SMC effect that transiently suppressed stent intimal formation, and when no longer present on day 28, there is an apparent "catch-up" phenomenon resulting in an increase in cell density and negation of the antiproliferative effect observed on day 3 in the HP1/2-treated rabbits.
In summary, these data highlight the early, abrogative effects of inhibiting the
4 integrin subunit poststent implantation but suggest that sustained inhibition of inflammatory cell involvement is required for the long-term attenuation of stent IA and likely prevention of ISR.
| ACKNOWLEDGEMENTS |
|---|
Received December 8, 2003; revised January 25, 2004; accepted February 5, 2004.
| REFERENCES |
|---|
|
|
|---|
4 integrins in vivo J. Clin. Invest. 94,1722-1728
4 integrin and ICAM-1 markedly attenuate macrophage homing to atherosclerotic plaques in apoE-deficient mice Circulation 97,75-81
4 integrin antibody is associated with less neoadventitial formation after balloon injury of porcine coronary arteries Can. J. Cardiol. 16,187-196
4 and vascular cell adhesion molecule-1 during smooth muscle cell differentiation Circ. Res. 80,159-169This article has been cited by other articles:
![]() |
C. Deiner, E. Shagdarsuren, P. L. Schwimmbeck, P. Rosenthal, C. Loddenkemper, U. Rauch, M. Pauschinger, R. Dietz, H.-P. Schultheiss, R. Dechend, et al. Nf-{kappa}b and AP-1 activation is associated with late lumen loss after porcine coronary angioplasty and antiproliferative beta-irradiation Cardiovasc Res, July 1, 2007; 75(1): 195 - 204. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Rippstein, M. K. Black, M. Boivin, J. P. Veinot, X. Ma, Y.-X. Chen, P. Human, P. Zilla, and E. R. O'Brien Comparison of Processing and Sectioning Methodologies for Arteries Containing Metallic Stents J. Histochem. Cytochem., June 1, 2006; 54(6): 673 - 681. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y.-X. Chen, X. Ma, S. Whitman, and E. R. O'Brien Novel Antiinflammatory Vascular Benefits of Systemic and Stent-Based Delivery of Ethylisopropylamiloride Circulation, December 14, 2004; 110(24): 3721 - 3726. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |