Fibrin(ogen) and von Willebrand Factor Deposition Are Associated With Intimal Thickening After Balloon Angioplasty of the Rabbit Carotid Artery
Abstract The aim of the study was to assess the contribution of thrombus incorporation into neointimal thickening in the rabbit carotid artery after deep vascular injury induced by balloon angioplasty compared with superficial vascular injury induced by a perivascular collar. Besides CD 31 (PECAM I), vimentin, α-smooth muscle actin, rabbit anti-macrophage monoclonal antibody and proliferating cell nuclear antigen, fibrin(ogen) and von Willebrand factor (vWF) deposition was assessed immunohistochemically. Angioplasty was performed in 47 rabbits and evaluated immediately (n=7), after 6 hours (n=4), and after 1 (n=7), 2 (n=9), or 3 (n=20) weeks. A collar was placed in 29 rabbits and evaluated immediately (n=5), after 6 hours (n=5), and after 1 (n=7), 2 (n=10), or 3 (n=2) weeks. After dilatation, the arteries were extensively denuded of endothelium, the internal elastic membrane was ruptured and blood-filled clefts were present in the media, pointing to deep vascular (type III) injury. Six hours later, mural fibrin(ogen) thrombi were formed, specially at sites with severe damage. This fibrin(ogen) matrix became infiltrated by phagocytes and smooth muscle cells. A luminal cap covered by regenerating endothelium was formed, demonstrating increased immunoreactivity to vWF. vWF was deposited in the extracellular neointimal spaces. Fibrin(ogen) thrombus deposition and incorporation appeared to be protracted phenomena for at least 2 weeks. After collar placement, minimal endothelial denudation was documented, pointing to focal superficial (type I) vascular injury. In subsequent weeks, neointimal thickening was associated with vWF deposition but not with fibrin(ogen) thrombus incorporation. In conclusion, mural fibrin(ogen) thrombus formation and incorporation contribute to neointima formation after deep vascular injury and seem to occur for several weeks after the initial insult.
- Received May 15, 1995.
- Accepted May 23, 1996.
PTCA has found widespread application in the treatment of different subsets of ischemic heart disease since its introduction in 1979 by Grüentzig et al.1 Although the immediate success rate of both balloon dilatation and other newly introduced interventional techniques has increased to more than 95%, restenosis during early follow-up remains a major limitation of coronary angioplasty. Clinical and quantitative angiographic follow-up studies have demonstrated that within 4 to 6 months of successful balloon dilatation, ≈30% of patients develop significant coronary artery restenosis.2 3 A wide variety of pharmacological therapies has been tested and various new technologies have been introduced, but no effective solution has yet been found for this important clinical problem.4 5 6
Although the precise mechanism remains unclear, restenosis is thought to result from excessive myointimal proliferation.7 8 9 Medial SMCs respond to dilatation injury by migration, proliferation, and extracellular matrix synthesis. The myointimal growth process has been attributed to release of chemoattractants and growth factors released by activated platelets that adhere to the site of injury of the vessel wall10 11 12 13 14 and cytokines released by other cells.7 8 Myointimal proliferation leads to the formation of a space-occupying neointima that progressively obstructs the arterial lumen, thereby leading to restenosis. Inhibition of platelet aggregation by aspirin,15 16 dipyridamole,16 ticlopidine,17 or anti-glycoprotein IIb/IIIa antibodies18 has reduced neointima formation in animal models. Nevertheless, the role of platelets as the primary “trigger” of restenosis after balloon angioplasty remains controversial, since all platelet-inhibitor intervention studies have failed to prevent human restenosis.19 20
Concepts about the pathogenesis of restenosis after angioplasty were mainly derived from animal models, in which endothelial denudation was the major stimulus of platelet aggregation and neointimal proliferation.10 11 12 13 14 21 22 Endothelial denudation in most of those models was obtained by passing a Fogarty balloon through the artery, a technique that creates a superficial (type I) vascular wall injury.8 10 13 14 Balloon dilatation in the clinical setting, however, frequently induces a more severe and deeper (type III) lesion in the vascular wall23 that also may lead to important mural thrombus deposition.
Therefore, the aim of the study was to analyze the time course and contribution of thrombus incorporation to intimal thickening induced by balloon angioplasty, which induces deep vascular (type III) injury, compared with the focal superficial injury induced by a perivascular collar,24 25 26 which fulfills the criteria8 of a type I injury. In view of this aim, anticoagulant therapies were omitted. The rabbit carotid artery was used as a model, and thrombi were evaluated by light microscopy and specific immunohistochemistry for fibrin(ogen). Other vascular components were identified by specific stains for α-SM actin, vWF, platelet-EC adhesion molecule 1 (CD 31, or PECAM I), a rabbit anti-macrophage monoclonal antibody (RAM 11), vimentin, and PCNA.
Male New Zealand White rabbits (2.5 to 3 kg, n=81) fed a normal laboratory diet without cholesterol supplementation were anesthetized with sodium pentobarbital (30 mg/kg body weight IV). The rabbits were posed in a supine position and the ventral neck region infiltrated with xylocaine for local anesthesia. Both carotid arteries were surgically exposed and dissected from the surrounding tissues.
For angioplasty, direct arteriotomy was performed in the right common carotid artery just proximal to the bifurcation. A standard 2-cm-long, 2.5-mm PTCA balloon dilatation catheter (Advanced Cardiovascular Systems) was introduced over a 0.014-in. floppy guide wire and advanced toward the aorta into the carotid artery to ≈5 cm from the incision. The balloon was inflated three times with an inflation pressure of 6 atm for a period of 2 minutes each. Between subsequent inflations the deflated balloon was left in the artery for 1 minute. After removal of the catheter, the small distal incision was surgically closed and arterial blood flow restored.
Histological evaluation of the angioplastied region (2 cm) was performed immediately afterward without restoration of arterial blood flow (n=7) and after 6 hours (n=4) and 1 (n=7), 2 (n=9), or 3 (n=20) weeks. To this end and to prevent postmortem thrombus formation, the rabbits were anticoagulated with heparin (150 U/kg IV) and anesthetized with sodium pentobarbital (30 mg/kg body weight IV). The angioplastied region in the right carotid artery and the equivalent normal control region in the left carotid artery were surgically removed and immediately immersed in methacarn fixative (60% methanol; 30% 1,1,1-trichloroethane; and 10% glacial acetic acid). To preserve fresh mural thrombi, the arteries were not flushed with saline. Perfusion fixation was not used. Afterward the rabbits were killed with an IV overdose of pentobarbital.
In 5 rabbits a sham operation was performed. After induction of intravenous anesthesia and dissection of both carotid arteries, direct arteriotomy was performed in the right carotid artery. Eight minutes later the incision was closed again without inserting a balloon catheter. Histological evaluation was performed after 2 weeks. All surgical procedures conformed to the guidelines detailed in the Position of the American Heart Association on Research Animal Use and approved by the institutional animal use committee.
For focal, superficial injury induced–intimal thickening, a nonocclusive, biologically inert, soft, flexible silicone collar was placed around the middle part of the left common carotid artery and closed with silicone glue. The right common carotid artery underwent sham operation; ie, it was separated from the surrounding connective tissue and vagus nerve and received a stretch similar to that applied to the contralateral collared artery.24 25 26 Histological evaluation was also performed immediately (n=5) and after 6 hours (n=5) and 1 (n=7), 2 (n=10), or 3 (n=2) weeks.
For light microscopy, the arteries were fixed in methacarn. Afterward the segments were cut to a length of 4 mm. At least 4 segments per artery were paraffin embedded. Transverse sections were stained with Sirius red–hematoxylin and immunohistochemistry was performed. The reactions were performed with the indirect peroxidase antibody conjugate technique. The following monoclonal antibodies were used: α-SM actin (Sigma Chemical Co, 1/3000 dilution), PCNA (PC 10 cyclin, 1/10 dilution), vimentin (1/200 dilution), rabbit anti-macrophage (RAM 11, 1/100 dilution), and CD 31 (1/10 dilution; all from Dako). The sections were preincubated with BSA to omit specific binding of the primary antibody. The monoclonal antibodies were diluted in PBS. After three washes in PBS, the sections were incubated with rabbit anti-mouse peroxidase (The Jackson Laboratory) for 45 minutes. For demonstration of the complex, 3-amino-9-ethylcarbazole was used as the chromogen. The specificity of these antibodies was demonstrated in previous studies.24 25 26
The following polyclonal antibodies (raised in sheep) were used: vWF (1/250, Binding Site) and rabbit fibrin(ogen) (1/10 000 dilution, Cappel). The specificity of the primary antibody against vWF was determined previously.25 The specificity of the primary antibody against fibrin(ogen) was tested by the staining pattern of prepared fibrin clots of rabbit plasma. The clots showed dense, threadlike, immunoreactive conglomerates that we considered to be fibrin. However, the antibody was not specific for fibrin because preabsorption of the antibody with fibrinogen had completely abolished immunoreactivity within the clots. This observation indicated that the antibody recognized both fibrin and fibrinogen. The term fibrin(ogen) is therefore used to indicate that the antibody does not distinguish fibrin from fibrinogen. The polyclonal antibodies were visualized by pig anti-sheep peroxidase as the secondary antibody (Binding Site), with 3-amino-9-ethylcarbazole as the chromogen or donkey anti-sheep fluorescein (The Jackson Laboratory) for double immunofluorescence. For double immunofluorescence both primary antibodies were used in parallel. For negative controls the primary antibody was omitted.
Quantification of Areas of vWF and Fibrin(ogen) Accumulation
The presence of fibrin(ogen) and vWF in the neointima or media was analyzed by a color image analyzing system (PC-image Colour, Foster Findlay Associates). Quantification was performed for at least 4 segments from each artery. Only for those arteries harvested immediately after angioplasty or collar placement was analysis not performed. The arterial wall of each segment was divided into four quadrants. In each quadrant the regions between the luminal margin and the IEL (neointima) and between the IEL and the EEL (media) were demarcated. The lengths of the IEL and EEL, the areas of neointima and media, and the vWF- and fibrin(ogen)-immunoreactive areas within both regions were measured. Segmentation was performed by measuring the average brown color of 10 different points within the immunoreactive areas. Percentages of vWF- and fibrin(ogen)-immunoreactive areas within the neointima and media were calculated.
Quantification of Intimal and Medial Thickening
The α-SM actin–stained sections were used to count the maximum number of SMC layers in the media and neointima (between the IEL and the luminal margin) in at least 4 segments from each artery. The mean of the maxima of these 4 segments was calculated.
Quadrant Injury Score
For each quadrant of arterial segment, a quadrant injury score was calculated (minimum of 0, maximum of 3), as defined by the sum of the presence (+1) or absence (0) of endothelium, IEL ruptures, and clefts in the media. Since all arteries were immersion fixed at ambient atmospheric pressure, the IEL always has a regular undulating aspect in normal rabbit carotid arteries. In contrast to those in human arteries, pores are normally small in the IEL of rabbit carotid arteries,27 28 29 and ruptures were defined as large irregularities in the IEL, which sometimes appeared as calcified discontinuities in regions of severe injury.
Presence of ECs
At least 4 segments from each artery were examined. Lumen-lining, nucleated cells that were immunoreactive for vimentin, demonstrated membrane-associated immunoreactivity for CD 31 and granular cytoplasmic immunoreactivity for vWF, and were negative for α-SM actin were considered to be ECs. Their presence was classified visually as absent, focal, or continuous.
Presence of Thrombus
A thrombus was identified as the luminal presence of amorphous material consisting of fibrin(ogen), RBCs, and platelets. The thrombus was regarded as organized if it had been infiltrated by cellular elements, such as SMCs and macrophages (RAM 11–positive cells), or if it had been covered by ECs. Thrombi could be occlusive or nonocclusive. Luminal threadlike, dense, fibrin(ogen)-immunoreactive conglomerates were considered to be fibrin. Platelets were defined as CD 31–positive, vWF-positive, vimentin-negative, nonnucleated, nonendothelial elements. The number of SMCs and PCNA-positive cells within each thrombus was counted and the thrombus area measured in at least 4 segments from each artery. The mean of the 4 segments was calculated and expressed as number per square millimeter.
All data are expressed as mean±SEM; n refers to the number of rabbits. The percentage of vWF- and fibrin(ogen)-immunoreactive area within the neointimal or medial region of interest and the extent of neointimal thickening of the collared and dilatated arteries were compared by the nonparametric Mann-Whitney U test. The spss for Windows package (SPSS) was used for these purposes. A 5% level of significance was selected.
Sham-operated arteries from both collared and dilatated animals were evaluated after 2 weeks. At this time an intact EC layer was present, resting directly on the intact IEL. The ECs demonstrated membrane-associated immunoreactivity for CD 31 and granular cytoplasmic immunoreactivity for vWF. Because fixation had been performed under ambient atmospheric pressure, the IEL always displayed a regular, continuous, undulating aspect in which pores were not visible. Mural fibrin(ogen) thrombi were never present. Subendothelial α-SM actin–positive cells were absent. In the media, defects were not seen and fibrin(ogen) deposition was absent. In every aspect these sham-operated segments were indistinguishable from naive, nonmanipulated rabbit carotid arteries (results not shown).
Numerical results are summarized in Tables 1⇓ and 2⇓. Immediately after angioplasty (Fig 1⇓), marked segmental loss of SMCs and deep, circular medial tears were always observed (7/7). Although ruptures of the IEL were visible in only 2 of 7 arteries (complete loss in one and ruptures in 3/4 quadrants in the other), focal, flattened distension of the IEL pointed to injury in all segments (Fig 1⇓). This distension, or focal loss of the undulating aspect, was never present in the contralateral control arteries. All arteries with remnants of an IEL(6/7) were partially covered by scarce “islands” of remaining ECs (Fig 3b⇓), which were identified as such by the presence of nuclei and immunoreactivity for vimentin, vWF, and CD 31 and the absence of immunoreactivity for α-SM actin.
Arteries With Occlusive Thrombi After Angioplasty
In 30% (n=12/40) of dilatated arteries an occlusive thrombus was observed. The 7 rabbits that were evaluated immediately after angioplasty were not included in this series because in this group, arterial blood flow had not been restored after dilatation. The injury of the IEL in arteries with occlusive thrombi (ruptures in 2.2±0.5 quadrants) was significantly more severe (P<.001, Mann-Whitney U test) than in arteries with nonocclusive thrombi (ruptures in 0.2±0.1 quadrant).
Six hours after dilatation, 1 of 4 examined arteries was completely occluded by a very loose luminal thrombus consisting of fibrin(ogen) and RBCs on top of a severely damaged IEL and media, also containing fibrin(ogen) deposits.
After 1 week, 2 of 7 dilatated arteries were occluded by a loose luminal thrombus, again consisting of fibrin(ogen) and RBCs. There was a marked influx of RAM 11–immunoreactive macrophages and a few α-SM actin–positive SMCs in the thrombus in both cases.
Two weeks after dilatation, 3 of 9 arteries showed occlusive fibrin(ogen) thrombi with more pronounced infiltration by macrophages (RAM 11–positive cells) and SMCs. Severe damage to the media with marked circular tears and fibrin(ogen) deposits was still visible. The intima remained unchanged: neither intimal SMCs nor ECs were observed.
After 3 weeks (Fig 2⇓), occlusive fibrin(ogen) thrombi (6/20) were colonized further by numerous macrophages and many SMCs (Fig 2a⇓), and collagen deposition had started. Most of the RAM 11–positive macrophages at this time were loaded with iron or hemosiderin (Fig 2b⇓), probably due to phagocytosis of RBCs. Clusters of foam cells and multinucleated giant cells (Fig 2c⇓) were sometimes seen. Neovascularization of the thrombus by several capillaries was also observed (Fig 2b⇓). Intimal thickening was not present. Damage of the media with circular clefts and fibrin(ogen) deposits was still always visible.
Arteries With Nonocclusive Thrombi After Angioplasty
Seventy percent (n=28/40) of arteries remained patent after balloon dilatation. Six hours after dilatation (Fig 3a⇓) thin, mural thrombi consisting of dense, threadlike conglomerates of fibrin(ogen) and RBCs were observed in all preparations (3/3), in close contact with sites of severe, deep injury of the media. In these 3 arteries (12 segments, 48 quadrants), 5 mural thrombi were recognized. The mean injury score of the quadrants covered by thrombi was significantly higher than that of quadrants not covered by thrombi (2.6±0.2 versus 1.3±0.3; P<.001, Mann-Whitney U test). Fibrin(ogen) deposits (Fig 3a⇓) and numerous RBCs (Fig 1⇑) were observed in deep medial clefts. Only scarce CD 31–positive, vWF-positive, vimentin-negative, nonnucleated platelets were observed in 1 of 3 specimens. The ruptured and extended internal elastic membrane was partially covered by islands of remaining ECs (Fig 3b⇓ and Table 2⇑).
One week after dilatation, in addition to a modest circular intimal thickening (on average, two SMC layers; Table 1⇑), mural thrombi were present in all 5 arteries (Fig 3c⇑) and again mainly consisted of fibrin(ogen). This fibrin(ogen) matrix was infiltrated by several α-SM actin–positive cells (Fig 3c⇑). Most of these SMCs had become oriented longitudinally, although in more bulging thrombus areas the organization pattern remained rather random. RAM 11–immunoreactive macrophages were present in all thrombi and in areas of medial injury. Fibrin(ogen) deposits could still be recognized in medial regions of injury. At this time, CD 31– and vWF-positive ECs were present in all specimens, predominantly in a focal way (4/5; Table 2⇑).
Two weeks after dilatation (Fig 4⇓), a circular, asymmetric intimal thickening with an average of five SMC layers (Table 2⇑) was observed in all 6 arteries. This thickened layer consisted of radially oriented SMCs above the IEL and longitudinally oriented cells on the luminal side (Fig 4d⇓). The ratio of intimal area to IEL length was 28±6 μm. In bulging intimal regions clearly due to incorporation of organizing mural fibrin(ogen) thrombi, the SMC orientation pattern was rather random (Fig 4d⇓). RAM 11–immunoreactive macrophages were again detected predominantly in the organizing thrombi and the remaining areas of medial injury. Throughout the entire intima a striking fibrin(ogen) immunoreactivity was present, sometimes most pronounced in the deeper parts in close contact with the internal elastic membrane (Fig 3f⇑). In all arteries, the intima had become lined with ECs, which had already formed a continuous layer in most areas (Figs 3e⇑ and 4e⇓). A relatively high fraction of EC nuclei stained positively for PCNA, suggestive of intense proliferative activity (Fig 4b⇓). The ECs showed dense, flocculent immunoreactivity for vWF, and vWF-immunoreactive material was found beneath the endothelium as well (Fig 3d⇑). The latter material was found to be extracellular, as demonstrated by double immunofluorescence for α-SM actin and vWF (results not shown). The depth of subendothelial accumulation varied. The media was nearly completely repaired. Some scarce medial fibrin(ogen) deposits could still be detected (Fig 3f⇑). In most arteries (5/6), the mural thrombi, which had become incorporated into the intima, demonstrated an unequal degree of organization (Fig 4a⇓). Some had been completely organized by SMCs and macrophages and showed incipient neovascularization. Others appeared to have formed only recently with a loose and amorphous aspect, in which infiltration and organization had just started. PCNA immunohistochemistry (Fig 4b⇓) demonstrated that the cells in these “younger”-looking thrombi showed a high cell replication rate, whereas the media and older-looking thrombi in the same specimen did not show PCNA-reactive nuclei, indicating a lower replication rate. At this time the media appeared to be almost completely repaired, and the number of SMC layers was clearly increased (on average, 13 SMC layers; Table 2⇑).
Three weeks after dilatation a prominent, circular, frequently eccentric intimal thickening had formed (14/14). In these eccentric intimal areas, almost completely organized mural thrombi could be recognized with significant collagen deposition, numerous SMCs with rather random orientation pattern, some macrophages, and in most cases (10/14) pronounced neovascularization with many small capillaries. An unequal degree of thrombus organization was sometimes still distinguishable (2/14). Again, striking fibrin(ogen) immunoreactivity was observed throughout the whole intima, sometimes still most pronounced in the deeper parts (Fig 3g⇑). ECs were always present, and the lining was often (10/14) continuous (Table 2⇑). A relatively high fraction of the EC nuclei still stained positive for PCNA (results not shown). The ECs’ dense, flocculent immunoreactivity for vWF and extracellular vWF immunoreactive material was found beneath the endothelium. The depth of subendothelial accumulation varied. Some scarce medial fibrin(ogen) deposits could still be detected (Fig 3g⇑). The number of SMC layers in the completely repaired media was still moderately increased (on average, 10 SMC layers; Table 2⇑).
Numerical results are summarized in Table 3⇓. Immediately after collar placement (n=5), only several small foci of EC denudation were present.25 The IEL and the media were without visible defect.24 25
Six hours after collar placement (n=5), the media showed segmental infiltration by polymorphonuclear cells and focal immunoreactivity for fibrin(ogen) (Fig 5a⇓). The fibrin(ogen) deposition was most pronounced in the inner third of the media, especially in regions with dense polymorphonuclear cell infiltration. vWF had not been deposited in the media. In contrast to the dilatated arteries, mural fibrin(ogen) thrombi were never observed and the intimal surface was not covered by fibrin(ogen) deposits. However, the periadventitia showed a slight concentric fibrinogen deposition, possibly related to surgical manipulation.
One week after collar placement (n=7), a small, luminal intimal thickening could be recognized consisting of a maximum of two to three layers of subendothelial SMCs. Mural fibrin(ogen) thrombi were absent, and fibrin(ogen) immunoreactivity could not be recognized in either intima or media. The EC layer was completely intact but demonstrated dense cytoplasmic immunoreactivity for vWF. Moreover, a variable diffuse immunoreactivity was found in the inner media.25
Two weeks after collar placement (n=10), a circular subendothelial accumulation of two to five cell layers of α-SM actin–positive SMCs (Table 3⇑) was present in all 10 rabbits. The continuous layer of CD 31–positive ECs (Fig 5d⇑) showed a dense flocculent immunoreactivity for vWF (Fig 5b⇑). Also beneath the endothelium extracellular vWF-immunoreactive material was found. Double immunofluorescence for α-SM actin and vWF demonstrated that the subendothelial vWF deposits lay in the extracellular space between the intimal SMCs (results not shown). The depth of subendothelial accumulation varied. In some regions focal immunoreactivity was found in the inner media. Mural fibrin(ogen)-rich thrombi were absent (Fig 5c⇑). Diffuse, intimal, fibrin(ogen)-positive material was seen in 1 specimen only, predominantly in subendothelial regions. The media never contained fibrin(ogen). The internal elastic membrane was intact. The ratio of intimal area to IEL length was 17±4 μm.
Three weeks after collar placement (n=2), a concentric neointima consisting of five to six cell layers was present. The medial SMCs appeared normal. Fibrin(ogen) deposition could not be demonstrated.
Quantification of the Areas of Fibrin(ogen) and vWF Immunoreactivity
At all evaluated time points, the mean area of fibrin(ogen)-immunoreactive material in the neointima and media was significantly higher in dilatated arteries than in collared arteries (Fig 6⇓). Moreover, in dilatated arteries a significant relation existed between quadrant injury score and the quadrant area of fibrin(ogen)-immunoreactive material in the neointima and media at all evaluated time points (Fig 7⇓). However, a significant linear correlation between quadrant area of fibrin(ogen)-immunoreactive material in the neointima and total quadrant area of the neointima after dilatation could not be demonstrated (1 week: r2=.078, NS; 2 weeks: r2=.132, NS; 3 weeks: r2=.019, NS; Fig 8⇓).
The mean vWF-immunoreactive area in the neointima of collared compared with dilatated arteries was not statistically different at all evaluated time points (Fig 9⇓).25 No relation existed between quadrant injury score and the quadrant area of vWF-immunoreactive material at all evaluated time points after dilatation. A significant linear correlation between quadrant area of vWF-immunoreactive material in the neointima and total quadrant area of the neointima after at different time points could not be demonstrated (1 week: r2=.132, NS; 2 weeks: r2=.174, NS; 3 weeks: r2=.098, NS).
Quantification of Mural Fibrin(ogen) Thrombus Organization Over Time
During the first 2 weeks after angioplasty, significant differences between the mean number of SMCs per unit area and the mean number of PCNA-positive cells per unit area in different mural thrombi from 1 artery could be demonstrated. From the third week on, no quantitative regional differences could be documented (Table 4⇓).
Choice of Models
To evaluate how the contribution of thrombus formation to intimal thickening could be influenced by the extent of injury, two completely different types of vascular injury were compared in the present study. The perivascular collar is one of the most gentle methods used to induce intimal thickening. In the early postoperative period, collared arteries demonstrate only small foci of endothelial denudation,24 25 which regenerate rapidly. However, there are some early changes in endothelial function, with impaired muscarinic receptor activity26 and increased vWF immunoreactivity.25 Moreover, experiments with radiolabeled fibrinogen point to an early increase in endothelial permeability.30 The latter finding was confirmed in the present study, with scarce subendothelial fibrin(ogen) deposition predominantly at sites of neutrophil infiltration. Because of the modest amount of morphological endothelial damage that it causes, the perivascular approach induces damage that can be classified as a focal type I vascular injury8 and is therefore related to the previously studied animal models.21 The lack of mural thrombus formation in the collar model is an additional similarity to the models of superficial vascular injury studied previously.21 On the contrary, inflation of a slightly oversized coronary angioplasty balloon catheter via a protocol that is similar to the one used during clinical coronary angioplasty created extensive vascular wall damage with intramural clefts. Since this model causes a deep type III injury, it appears to be more representative of clinical balloon angioplasty.23
Mural Fibrin(ogen) Thrombi and Intimal Thickening
After collar placement, neointimal thickening is not associated with mural fibrin(ogen) thrombus incorporation despite the presence of scarce early, transient medial fibrin(ogen) deposits. The latter are possibly related to the increased permeability of the EC layer, which is also indicated by the increased transarterial fibrinogen flux in collared arteries.30
Immediately after balloon dilatation the arteries were extensively denuded of endothelium, and deep intramural hemorrhages within the media, sometimes extending to the EEL, were present. Six hours later dense, threadlike conglomerates of fibrin(ogen)-immunoreactive material were present in both mural thrombi and the medial layer. The fibrin(ogen) matrix became incorporated into the neointimal thickening by extensive phagocytic invasion, infiltration by SMCs, and formation of a luminal cap covered by regenerating endothelium. This process was accompanied by neovascularization, which provides further proof for the important contribution of fibrin(ogen) thrombi.31 The neointimal fibrin(ogen) matrix could be detected as late as 3 weeks after angioplasty. A portion of this fibrin(ogen), in closer contact with the IEL, may be a remnant of the initial deposit. Another portion of this fibrin(ogen), however, may be due to increased permeability of the rapidly regenerating endothelium,30 with leakage of fibrinogen from the arterial lumen through the endothelium into the neointimal tissue and local secondary conversion to fibrin.32 33 34 35 36 37
Mural fibrin(ogen) thrombus deposition and incorporation seem to be protracted phenomena that are not limited to the initial trauma. Mural thrombi with different degrees of organization were frequently demonstrated within 1 segment. Some were completely organized by SMCs and macrophages and showed incipient neovascularization. Others appeared to be only recently formed with a loose, amorphous aspect in which infiltration and organization had just started. Quantitative PCNA immunohistochemistry (Table 4⇑) demonstrated that particularly in younger-looking thrombi, the invading cells (SMCs and monocytes) were actively replicating, while in the adjacent, older-looking thrombi the cell replication rate was much lower. Although real age differences can not be proved by our methodology, the striking differences between adjacent thrombi within 1 segment clearly suggest differences in cell proliferation rate and age. This continuous mural thrombosis is possibly related to the areas with marked endothelial denudation, which persist throughout the first 2 weeks. Afterward, when the endothelium has almost completely regenerated, new mural fibrin(ogen) thrombi are rarely formed. This element, which to our knowledge is a new finding, has not been mentioned in other studies and could have implications for the continuation of anticoagulant therapy after clinical PTCA.
One third of these fibrin(ogen) thrombi became occlusive. As we have demonstrated, mural fibrin(ogen) thrombi are mostly formed in quadrants with a high injury score (Fig 7⇑). Moreover, IEL injury in occluded arteries was significantly more severe than that in nonoccluded arteries. Therefore, the thrombotic process seems to be related to the severity of the vascular insult. The occlusive thrombi can probably be considered the experimental equivalent of the acute thrombotic occlusions that sometimes occur after PTCA in humans. The higher incidence of acute thrombosis in our model can be partially attributed to the absence of any antiplatelet or anticoagulant therapy before, during, or after balloon dilatation.
On the basis of findings for the perivascular collar and balloon angioplasty models, we have demonstrated that intimal thickening can be considered a general reaction pattern induced by vascular injury but is produced by different mechanisms depending on the type or severity of injury. Fibrin(ogen) accumulation is certainly not an essential element for intimal induction, but it appears to contribute to neointimal thickening after type III injury. However, no clear linear mathematical correlation was found between intimal thickening and fibrin(ogen) immunoreactivity; this lack is presumably due to fibrinolysis. Indeed, the amount of fibrin(ogen) immunoreactivity decreased in both the intima and media as time progressed (Fig 6⇑). To prove the contribution of fibrin(ogen) deposition to restenosis more conclusively, further experiments are necessary in which type III injury is combined with either the blockade of fibrin and/or thrombin formation or the stimulation of fibrinolysis. On the basis of results presented herein, studies based on type I injury, like the frequently used Fogarty balloon denudation model, can give misleading information with respect to PTCA in the clinical situation.
Intimal Thickening and vWF
The increased flocculent immunoreactivity for vWF in regenerating ECs after dilatation and the presence of vWF in the subendothelial extracellular spaces parallel the findings from the collar model.25 In normal arteries ECs secrete vWF into the plasma and toward the subendothelial space, where minimal deposition of this glycoadhesive protein is present in the basal lamina.38 Some authors attribute an important function in EC adhesion to the basal lamina vWF.39 40 41 In cell culture studies it has been demonstrated that ECs can be stimulated to release additional large amounts of vWF under conditions of stress.40 The effect of the increased subendothelial vWF deposition on neointimal formation remains speculative but might by itself contribute to SMC migration and neointimal formation.25
In conclusion, in the rabbit carotid artery, balloon dilatation created a type III injury of the vascular wall, resulting in the formation of mural fibrin(ogen) thrombi that became progressively incorporated into the intima. The thrombotic process was related to the severity of the injury and appeared to proceed for weeks. On the contrary, intimal thickening after collar placement was not associated with fibrin(ogen) accumulation. Although vascular organization of thrombi has been known for years, the present study suggests that mural thrombus deposition and its incorporation into the intima play a significant role in restenosis after coronary angioplasty. Long-term prevention of continuing mural fibrin(ogen) thrombus deposition may therefore offer an effective therapeutic strategy for the prevention of restenosis.
Selected Abbreviations and Acronyms
|EEL||=||external elastic lamina|
|IEL||=||internal elastic lamina|
|PCNA||=||proliferating cell nuclear antigen|
|PTCA||=||percutaneous transluminal coronary angioplasty|
|RBC(s)||=||red blood cell(s)|
|SMC(s)||=||smooth muscle cell(s)|
|vWF||=||von Willebrand factor|
Serruys PW, Luiten HE, Beatt KJ, Geuskens R, De Feyter PJ, Van den Brand M, Reiber JHC, Ten Katen HJ, van Es GA, Hugenholtz PG. Incidence of restenosis after successful coronary angioplasty: a time-related phenomenon—a quantitative angiographic study in 342 consecutive patients at 1, 2, 3 and 4 months. Circulation. 1988;77:361-371.
Kuntz RE, Baim DS. Defining coronary restenosis. newer clinical and angiographic paradigms. Circulation. 1993;88:1310-1323.
Topol EJ, Leya F, Pinkerton CA, Whitlow PL, Hofling B, Simonton CA, Masden RR, Serruys PW, Leon MB, Williams DO, King-SB I, Mark DB, Isner JM, Holmes DH Jr, Ellis SG, Lee KL, Keeler GP, Berdan LG, Hinohara T. A comparison of directional atherectomy with coronary angioplasty in patients with coronary artery disease. N Engl J Med. 1993;329:221-227.
Liu MW, Roubin GS, King SB III. Restenosis after coronary angioplasty: potential biological determinants and role of intimal hyperplasia. Circulation. 1989;79:1374-1387.
Fingerle J, Johnson R, Clowes AW, Majesky MW, Reidy MA. Role of platelets in smooth muscle cell proliferation and migration after vascular injury in rat carotid artery. Proc Natl Acad Sci U S A. 1989;86:8412-8416.
Grotendorst GR, Seppa HEJ, Kleinman HK. Attachment of smooth muscle cells of collagen and their migration toward platelet-derived growth factor. Proc Natl Acad Sci U S A. 1981;78:3669-3672.
Bernstein LR, Antantoniades H, Zetter BR. Migration of cultured vascular cells in response to plasma and platelet-derived factors. J Cell Sci. 1982;56:71-82.
Lindner V, Fingerle J, Reidy MA. Mouse model of arterial injury. Circ Res. 1993;73:792-796.
Volker W, Faber V. Aspirin reduces the growth of medial and neointimal thickenings in balloon-injured rat carotid arteries. Stroke. 1990;21(suppl IV):IV-44-IV-45.
Badimon L, Badimon JJ, Cohen M, Chesebro JH, Fuster V. Thrombus formation in stenotic and laminar flow conditions: effect of an anti-platelet GPIIb/IIIa monoclonal antibody fragment. Circulation. 1989;80(suppl I):I-422. Abstract.
Popma JJ, Califf RM, Topol EJ. Clinical trials of restenosis after coronary angioplasty. Circulation. 1992;84:1426-1436.
Haudenschild CC. Restenosis: pathophysiologic considerations. In: Topol EJ, ed. Textbook of Interventional Cardiology. Philadelphia, Pa: WB Saunders Co; 1994:382-399.
Kockx MM, De Meyer GRY, Jacob WA, Bult H, Herman AG. Triphasic sequence of neointimal formation in the cuffed carotid artery of the rabbit. Arterioscler Thromb. 1992;12:1447-1457.
Kockx MM, De Meyer GRY, Andries LJ, Bult H, Jacob WA, Herman AG. The endothelium during cuff-induced neointima formation in the rabbit carotid artery. Arterioscler Thromb. 1993;13:1874-1884.
De Meyer GRY, Bult H, Van Hoydonck A, Jordaens FH, Buyssens N, Herman AG. Neointima formation impairs endothelial muscarinic receptors while enhancing prostacyclin mediated responses in the rabbit carotid artery. Circ Res. 1991;68:1669-1680.
Sims FH. Discontinuities in the internal elastic lamina: a comparison of coronary and internal mammary arteries. Artery. 1985;122:127-143.
Coleman PJ, Lever MJ, Martin JF. Transarterial fibrinogen flux in the anaesthetized rabbit. J Physiol. 1993;467:41P. Abstract.
Houdijk WPM, de Groot PG, Nievelstein PFEM, Sakariassen KS, Sixma JJ. Subendothelial proteins and platelet adhesion: von Willebrand factor and fibronectin, not thrombospondin, are involved in platelet adhesion to extracellular matrix of human vascular endothelial cells. Arteriosclerosis. 1986;6:24-33.
Sporn LA, Marder VJ, Wagner DD. Differing polarity of the constitutive and regulated secretory pathways for von Willebrand factor in endothelial cells. J Cell Biol. 1989;108:1283-1289.
Dejana E, Lampugnani MG, Giorgi M, Gaboli M, Federici AB, Ruggeri ZM, Marchisio PC. von Willebrand factor promotes endothelial cell adhesion via an Arg-Gly-Asp dependent mechanism. J Cell Biol. 1989;109:367-375.