The relation among vessel injury, thrombus formation, and platelet survival in rabbits

Continuous or repeated injury of rabbit aortae by indwelling vascular catheters caused the deposition of platelets on the injured vessels and the formation of thrombi rich in platelets and fibrin at sites where flow was most disturbed and injury was most extensive. Incorporation of 51Cr platelets into the thrombi reached a maximum between 3 and 24 hr. The platelet-fibrin-rich thrombi remained reactive to circulating platelets for at least 14 days. Continuing reactivity of thrombi and the turnover of platelets in the thrombi were accompanied by an increase in the proportion of platelets that separated in the least dense fraction on Stractan density gradients. Platelet survival was also shortened (43.5 +/- 5.9 hr in animals with catheters, compared with 62.6 +/- 4.5 hr in animals with a sham operation), indicating that some platelets that had taken part in thrombus formation or had interacted with the injured vessel wall were rapidly cleared from the circulation. Platelets from rabbits that had had indwelling aortic catheters in place for 3 or 6 days survived significantly longer than those from animals with a sham operation upon injection of the platelets into normal animals; thus, continuous turnover of platelets on injured vessels and thrombi, and the clearance of altered platelets, leads to a population of younger platelets that survive longer. The continuing reactivity of thrombi may in part account for repeated occlusive episodes in vascular disease. The contribution of thrombin generation and fibrin formation to the platelet-rich thrombi is substantial and warrants the ongoing evaluation of treatment with a combination of anticoagulant and antiplatelet agents in arterial thrombosis and in thrombus formation on vascular catheters.     

Acute administration of ethanol to rabbits inhibits thrombus formation induced by indwelling aortic catheters

Platelet aggregation and secretion of granule contents in response to specific agonists are inhibited by acute ethanol (1 to 4 mg/ml in vitro or 4 to 5 mg/ml ex vivo). However, acute administration of ethanol, giving blood levels of 4 to 5 mg/ml, does not affect platelet adherence to de-endothelialized rabbit aortae in vivo, which is the initial step in thrombus formation. Whether ethanol inhibits subsequent reactions in thrombus formation induced by indwelling catheters in rabbit aortae was investigated. Before insertion of the catheters, the rabbits received injections of 51Cr-labeled platelets (and, in some experiments, [125I]fibrinogen), and were given ethanol by stomach tube, to achieve blood levels of approximately 4 mg/ml. Thrombus formation after 3 hours was assessed by determining the number of platelets and the amount of radiolabeled fibrin(ogen) associated with the aortae, and by determining the thrombus weights; all three measurements indicated a 64 to 84% reduction in thrombosis. Thus, experimentally induced thrombus formation in the aorta is significantly reduced by ethanol.     

Platelet accumulation and turnover on de-endothelialized aortae in rats

Previous studies indicate that the subendothelium of rabbit aortae de-endothelialized with a balloon catheter rapidly becomes covered with a monolayer of platelets; after 60 min few additional platelets accumulate and although most platelets are lost from the injured surface by 4 days, there is a substantial delay before re-endothelialization. We examined the dynamics of platelet accumulation on rat aortae de-endothelialized with a balloon catheter to determine if the response to this type of injury is similar to rabbit aortae. When 51Cr-platelets were injected prior to aortic de-endothelialization, 25,500 +/- 2,750 platelets/mm2 accumulated on rat subendothelium in the first 15 min. After 60 and 92 h, fewer platelets remained on the surface (13,740 +/- 2,400 and 5,020 +/- 1,330 platelets/mm2, respectively). When 51Cr-platelets were injected into rats 30 min after injury, platelet accumulation in a 30-min period was 8,610 +/- 1,230 platelets/mm2. By 4 days rat aortae did not accumulate newly injected platelets significantly in a 30-min period, but in a 24-h period 20,600 +/- 3,490 platelets/mm2 accumulated. Morphologically, the non-endothelialized areas of rat aortae were almost completely covered with platelets 4 days after injury. Fourteen days after injury, rat aortae did not accumulate newly injected platelets and, morphologically, no platelets were present on the surface which was almost re-endothelialized. Thus, in rats, as with rabbits, platelets rapidly accumulate on de-endothelialized aortae and the ability to attract newly introduced platelets is considerably reduced shortly after injury. In contrast to rabbits, however, de-endothelialized aortae in rats remain attractive to new platelets up to 4 days following injury, but less so than at the time of injury. Also, in contrast to rabbits, 14 days after injury to rat aortae the surface is almost completely re-endothelialized. Thus, there are species differences in platelet interactions with de-endothelialized vessels.     

Platelet accumulation and turnover on de-endothelialized aortae in rats.

Previous studies indicate that the subendothelium of rabbit aortae de-endothelialized with a balloon catheter rapidly becomes covered with a monolayer of platelets; after 60 min few additional platelets accumulate and although most platelets are lost from the injured surface by 4 days, there is a substantial delay before re-endothelialization. We examined the dynamics of platelet accumulation on rat aortae de-endothelialized with a balloon catheter to determine if the response to this type of injury is similar to rabbit aortae. When 51Cr-platelets were injected prior to aortic de-endothelialization, 25,500 +/- 2,750 platelets/mm2 accumulated on rat subendothelium in the first 15 min. After 60 and 92 h, fewer platelets remained on the surface (13,740 +/- 2,400 and 5,020 +/- 1,330 platelets/mm2, respectively). When 51Cr-platelets were injected into rats 30 min after injury, platelet accumulation in a 30-min period was 8,610 +/- 1,230 platelets/mm2. By 4 days rat aortae did not accumulate newly injected platelets significantly in a 30-min period, but in a 24-h period 20,600 +/- 3,490 platelets/mm2 accumulated. Morphologically, the non-endothelialized areas of rat aortae were almost completely covered with platelets 4 days after injury. Fourteen days after injury, rat aortae did not accumulate newly injected platelets and, morphologically, no platelets were present on the surface which was almost re-endothelialized. Thus, in rats, as with rabbits, platelets rapidly accumulate on de-endothelialized aortae and the ability to attract newly introduced platelets is considerably reduced shortly after injury. In contrast to rabbits, however, de-endothelialized aortae in rats remain attractive to new platelets up to 4 days following injury, but less so than at the time of injury. Also, in contrast to rabbits, 14 days after injury to rat aortae the surface is almost completely re-endothelialized. Thus, there are species differences in platelet interactions with de-endothelialized vessels.     

Distribution of cellular responses in rabbit aortae following one and two injuries with a balloon catheter.

To investigate the mechanisms involved in the cellular reactions to arterial injuries, we studied the distribution of the deposits on the injured intima and the pattern of neointimal thickening following single and double injuries of rabbit aortae with a balloon catheter. Thirty minutes after the first injury most, but not all, of the inner surface of the aortae was covered by adherent, spread platelets. Seven days following the first injury areas of neointima, mainly proliferating smooth muscle cells, had formed around and opposite the orifices of branch vessels. The rest of the inner aortic surface consisted of acellular subendothelial matrix. Thirty minutes after the second injury, 7 days after the first, single platelets adhered once more to parts of the reinjured subendothelium, mostly between the orifices. Numerous fibrin-rich, platelet thrombi were present mainly on the surface of the injured neointima. Thirty minutes after both the first and second injury polymorphonuclear leucocytes adhered to the inner surface downstream from the orifices of branch vessels and in longitudinally oriented zones opposite the orifices. Four days following the second injury, the neointima was restored with the same distribution as before the second injury, and few thrombi, adherent platelets, or leucocytes remained.     

Distribution of cellular responses in rabbit aortae following one and two injuries with a balloon catheter

To investigate the mechanisms involved in the cellular reactions to arterial injuries, we studied the distribution of the deposits on the injured intima and the pattern of neointimal thickening following single and double injuries of rabbit aortae with a balloon catheter. Thirty minutes after the first injury most, but not all, of the inner surface of the aortae was covered by adherent, spread platelets. Seven days following the first injury areas of neointima, mainly proliferating smooth muscle cells, had formed around and opposite the orifices of branch vessels. The rest of the inner aortic surface consisted of acellular subendothelial matrix. Thirty minutes after the second injury, 7 days after the first, single platelets adhered once more to parts of the reinjured subendothelium, mostly between the orifices. Numerous fibrin-rich, platelet thrombi were present mainly on the surface of the injured neointima. Thirty minutes after both the first and second injury polymorphonuclear leucocytes adhered to the inner surface downstream from the orifices of branch vessels and in longitudinally oriented zones opposite the orifices. Four days following the second injury, the neointima was restored with the same distribution as before the second injury, and few thrombi, adherent platelets, or leucocytes remained.     

Platelet interaction with damaged rabbit aorta.

We have quantified the accumulation of 51Cr-labeled washed rabbit platelets on the subendothelium of rabbit aortae following injury with a balloon catheter. The amount of radioactivity that became associated with the damaged wall within 10 minutes of the injury did not change appreciably during the following 24 hours, indicating that there was little turnover of platelets on the injured vessel wall. In addition, by injecting 51Cr-labeled platelets into rabbits at different times after injury, it was possible to estimate the reactivity of the exposed surface to newly injected platelets. Scanning electron microscopy showed that a monolayer of platelets initially formed on the injured surface; the number of platelets associated with the surface decreased over the 7-day observation period. The amount of 51Cr associated with the injured vessel wall also diminished during this period. The ability of the damaged surface to attract fresh platelets gradually decreased during the 7 days following injury. Platelet survival in rabbits was not significantly reduced following the removal of the aortic endothelium (balloon catheter injury 66.3 +/- 12.2 hours, sham operated 72.1 +/- 7.2 hours, untreated controls 76.2 +/- 3.8 hours). Thus, in rabbits, it cannot be assumed that platelet survival provides an estimate of endothelial injury in all circumstances.     

Real-time analysis of platelet aggregation and procoagulant activity during thrombus formation in vivo

The exact mechanism of blood vessel thrombus formation remains to be defined. Here, we introduce a new approach to probe thrombus formation in blood vessels of living animals using intravital microscopy in green fluorescent protein (GFP)-transgenic mice to simultaneously monitor platelet aggregation and procoagulant activity. To this end, GFP-expressing platelets and annexin A5 labeled with a fluorescent dye were employed to visualize and analyze platelet aggregation and markers of procoagulant activity (platelet surface phosphatidylserine (PS)). Laser-induced thrombi increased and then decreased in size with time in vessels of living animals, whereas platelet surface PS initiated at the site of injury and then penetrated into the thrombus. PS-positive platelets were predominantly localized in the center of the thrombus, as was fibrin generation. The experimental system proposed here is a valuable tool not only for investigating mechanisms of thrombus formation but also to assess the efficacy of antithrombotic drugs within the vasculature.     

Localization of Ferritin-Conjugated Anti-Fibrin/Fibrinogen in Platelet Aggregates Produced In Vitro

A ferritin-conjugated anti-fibrin/fibrinogen was localized by means of light and electron microscopy in artificial in vitro thrombi formed in the presence of the labeled antibody, and in preformed ADP-induced platelet aggregates. The ferritin was distributed throughout the central and peripheral regions of the columns of aggregated platelets in the thrombi. In the preformed ADP aggregates, ferritin was deposited only by infiltration from surrounding plasma and was confined to the periphery of the columns. The even distribution of ferritin in the central zone of the platelet columns of the thrombi indicated a specific reaction had occurred before or during thrombus formation unrelated to infiltration of plasma. In the artificial thrombi, the ferritin-labeled antibody was localized on the surface layer of platelets and on the bridging structures composed of the combined surface layers in the narrow spaces between cohesive platelets. Vesicles and alpha granules within the platelets also were tagged. The absence of obvious fibrin between narrow interspaces and within the platelets indicated that the antibody had reacted with fibrinogen or partly polymerized fibrin at these sites. Many invaginations of the platelet membrane containing dense fibrillar material were interpreted to be alpha granules discharging their contents during the “release reaction” at the time of aggregation. This material, which was tagged by the ferritin-conjugated antibody, merged with the interplatelet bridges to suggest that released fibrinogen from within the platelet contributed to the structural bond and strengthened it. A layer of dense fibrin and altered platelets in the periphery of the columns of aggregated platelets in the artificial thrombi contained the platelets and limited further growth of the aggregates. The fibrin was thought to be derived from infiltrated plasma as well as from released intraplatelet fibrinogen. Platelet fibrinogen thus appeared to take part both in the cohesion of aggregated platelets and in the stabilization of the aggregates formed.     

Effect of remote aortic injury and thrombosis on cholesterol atherosclerosis

Recent tissue culture studies indicate that platelet factors enhance lipoproteins to stimulate proliferation of smooth muscle cells, a key event in atherogenesis. It was of interest to determine if continued remote aortic injury and thrombosis can increase the severity of hypercholesterolemia induced atherosclerosis. Four groups of rabbits were studied, two groups fed 0.5 g cholesterol per day, one group with and one group without massive abdominal aortic white mural non-occlusive thrombosis and injury induced with a permanent indwelling catheter, and two groups fed 1 g cholesterol per day, one group with and one group without similar thrombosis and injury. Serum lipids in the four groups were qualitatively and quantitatively not significantly different from each other at 4, 8, and 12 weeks. When they were sacrificed at 12 weeks the percent intimal surface area in the thoracic aorta (48 ± 22) and arch (75 ± 28) of the rabbits with thrombosis and fed 0.5 g cholesterol per day was significantly (p < 0.01) greater than the percent intimal area (20 ± 14) and (24 ± 18) of the rabbits without thrombosis. The percent intimal area with lesions in the rabbits fed 1 g cholesterol per day with thrombosis (59 ± 25) and (84 ± 30) was significantly greater (p < 0.01) than the percent area (24 ± 29) and (32 ± 32) of the rabbits without thrombosis. Since the effects could not be attributed to differing serum lipid levels it is possible that the increased severity of atherosclerosis in the remote arch and thoracic aorta was related to increased permeability, cell proliferation or collagen synthesis possibly stimulated by circulating factors released from the remote massive non-occlusive thrombus or from circulating platelets activated by contact with the injured abdominal aortic wall.     

Formation and regulation of platelet and fibrin hemostatic plug

Formation of a hemostatic plug represents one of the earliest responses to vessel wall injury. Platelets react to any discontinuity in the vascular endothelium through initial contact, spreading, and formation of a thrombus (or aggregate). This development of a primary hemostatic plug requires platelet membrane receptors through which the adhesive macromolecules, von Willebrand factor (vWF) and fibrinogen, anchor platelets to the vessel wall and link them to each other. There are two receptor pathways--classic and alternative--for the binding of vWF to platelets; the latter induced by thrombin, and adenosine diphosphate (ADP) is shared with fibrinogen. Synthetic peptides, patterned after known binding domains of adhesive molecules, have been designed to inhibit their interactions with platelet receptors. A secondary hemostatic plug, composed of platelets enmeshed in fibrin, results from the action of thrombin, which is not only essential for formation of fibrin but also for exposure of platelet receptors for adhesive molecules and for "activation" of factors V and VIII. Thrombin generation is greatly enhanced through the activity of the prothrombinase complex formed on the surface of platelets, perturbed endothelial cells, and leukocytes. A pivotal event is activation of factor X through the intrinsic and extrinsic coagulation pathways. Binding of factors IXa and VIIa to the vascular endothelium represents a localized mechanism for factor Xa generation. Formation of a platelet and fibrin thrombus is controlled by regulatory mechanism: prostacyclin, endogenous heparin-antithrombin III complex, thrombomodulin-protein C-protein S system, and the fibrinolytic system. The balance of all components--vessel wall, platelets, adhesive and coagulation proteins, regulatory mechanisms--determines the effectiveness of the hemostatic plug in maintaining the structural and functional integrity of the circulatory system. An approach to detection of hemostatic derangements in patients at risk evolves from a full understanding of inherited and acquired deficiencies affecting each step of hemostatic plug formation and from selective use of laboratory tests.     

Platelet receptor signaling in thrombus formation

Platelet activation and subsequent thrombus formation at sites of vascular injury is crucial for normal hemostasis, but it can also cause myocardial infarction and stroke. The initial capture of flowing platelets to the injured vessel wall is mediated by the interaction of the glycoprotein (GP) Ib-V-IX complex with von Willebrand factor immobilized on the exposed subendothelial extracellular matrix. Tethered platelets are then able to bind to collagens through the immunoglobulin-like receptor GPVI and to initiate cellular activation, a process that is reinforced by G protein-coupled receptors stimulated by locally produced thrombin and soluble mediators released from activated platelets. These signaling events lead to a rise in the cytosolic Ca²⁺ concentration, rearrangement of the cytoskeleton, release of granule content, and functional upregulation of integrin adhesion receptors allowing firm adhesion and thrombus growth. Fully activated platelets also undergo a procoagulant conversion thereby facilitating coagulation and thrombus stabilization. This review summarizes the most important receptor systems and signaling mechanisms involved in platelet activation and thrombus formation with special focus on recent discoveries.     

The structure and growth of valve-pocket thrombi in femoral veins

The structure of 50 small thrombi in femoral valve pockets and the microscopic contents of 35 apparently empty pockets were studied in an attempt to ascertain the nature of the microscopic nidi from which thrombi form and their manner of growth to visible thrombi. Sixteen thrombi had little or no cellular invasion. Most of these recent structures had two main regions, red areas restricted distally in the pocket by the vein wall, and larger white regions comprising most of the thrombus length and often covering the red areas. Red areas are the early sites of cellular adhesion and invasion and the likely sites of origin of most thrombi. They were usually dominated by red cells and fibrin. White zones, which represent propagation growth, are characterized by many foci of platelets with fibrin borders (platelet-fibrin units). Some red areas also contained platelet-fibrin units but they were few and tiny; platelets were not seen in others and one small wholly red thrombus was devoid of platelets. Degenerative changes in platelet-fibrin units were observed, and it is postulated that many become purely fibrin structures. There was no significant evidence of preceding intimal damage in the vein wall. Therefore nidi are laid down on normal endothelium probably on the vein wall near the apex of the pocket. Some pockets, empty of thrombi, contained condensed foci of red cells or tiny fibrin fragments surfaced by endothelial cells and considered to be the remnants of aborted thrombi; a few contained clumps of platelets or leucocytes. It is postulated that any of these may represent the nidi from which thrombi grow. Several thrombi also incorporated large fat droplets, numerous in two. Fat embolic globules derived from fractures are their likely source.     

Thrombosis induced in vivo in the mesenteric artery of normal and thrombocytopenic rats, an electron-microscopic study of the early arterial wall reaction

Arteries taken from 1 to 3 days after local thrombosis had been induced by the passage of a weak electric current followed by the topical administration of ADP were studied by electronmicroscopy. The characteristic changes observed during this period included total disappearance of pre-existing thrombi, rapid reconstitution of the endothelium and important hypertrophy of smooth muscle cells in the media. Thrombocytopenia induced by the administration of antiplatelet serum produced no alterations in untreated arteries. When it was coupled to thrombus induction, healing of injured endothelial and smooth muscle cells was grossly impaired, hypertrophy of the media was absent and fibrin infiltrated the arterial wall at the site of thrombus induction. These results bring support to the view that blood platelets play an important role in the induction of the regeneration response of injured arterial cells.     

Thrombosis induced in vivo in the mesenteric artery of normal and thrombocytopenic rats,an electron-microscopic study of the early arterial wall reaction.

Arteries taken from 1 to 3 days after local thrombosis had been induced by the passage of a weak electric current followed by the topical administration of ADP were studied by electronmicroscopy. The characteristic changes observed during this period included total disappearance of pre-existing thrombi, rapid reconstitution of the endothelium and important hypertrophy of smooth muscle cells in the media. Thrombocytopenia induced by the administration of antiplatelet serum produced no alterations in untreated arteries. When it was coupled to thrombus induction, healing of injured endothelial and smooth muscle cells was grossly impaired, hypertrophy of the media was absent and fibrin infiltrated the arterial wall at the site of thrombus induction. These results bring support to the view that blood platelets play an important role in the induction of the regeneration response of injured arterial cells.     

Endogenously elicited antibodies to platelet derived growth factor-BB and platelet cytosolic protein inhibit aortic lesion development in the cholesterol-fed rabbit

Several studies have indicated that growth factors, such as platelet derived growth factor (PDGF), may be important in atherogenesis. These factors are released from platelets, or expressed by cells of the arterial wall. In order to study their role in atherogenesis more directly, rabbits were immunized with PDGF-BB, platelet cytosolic protein, or human serum albumin (HSA), until high titres of antibody were attained. Atherosclerotic lesions were subsequently induced by feeding the animals with a 2% cholesterol enriched diet. At the end of approximately 3 months, the extent of aortic lesion development was assessed by image analysis of en face preparations of aortae stained with Oil Red-O, and histological segments of aortae taken at the level of the first intercostal artery branch point. The endogenous antibodies were characterized with respect to their cross-reactivity, and ability to neutralize PDGF and platelet cytosol-induced cell proliferation and migration in vitro . The endogenous, anti-PDGF-BB antibody was isoform specific, and neutralized the mitogenic and chemotactic properties of PDGF-BB and rabbit platelet cytosolic protein in vitro . The anti-platelet cytosol antibody partially inhibited the chemotactic and mitogenic properties of rabbit platelet cytosolic protein. Compared to non-immune rabbits ( n =5), animals immunized with HSA ( n =4) had a significantly larger area of aortic lesion involvement ( P <0.01), whereas aortic lesions in rabbits immunized with PDGF-BB ( n =5), or platelet cytosolic protein ( n =7) were significantly smaller than either non-immune animals, or animals immunized with HSA ( P <0.05). The same pattern was observed for other measures of aortic lesion involvement including aortic intima : media ratio at the level of the first intercostal artery. These data suggest that PDGF-BB, and possibly other platelet-associated growth factors, are involved in cholesterol-induced atherosclerosis.     

Experimental meningococcal septicemia. Effect of aspirin therapy

In previous studies we have presented morphological evidence that the terminal shock-like phase of fatal meningococcemia is caused by the occlusion of the pulmonary microcirculation with thrombi composed of platelets, leukocytes, and fibrin. We have also shown that in experimental meningococcemia, pretreatment of rabbits with heparin sodium prevents fibrin formation but does not influence the cellular pulmonary thrombi and does not prolong survival. If our theory is correct, drugs that inhibit platelet aggregation and leukocyte adhesion in rabbits should prolong life. The present experiment demonstrates that pretreatment with a small dose of aspirin doubles the survival time without altering the mortality.     

Visualization and analysis of mural thrombogenesis on collagen, polyurethane and nylon

Visualization and analysis of mural thrombogenesis on collagen-coated glass, polyurethane and nylon was discussed. Epi-fluorescent video microscopy was used to visualize thrombotic events at a protein or polymer surface in contact with flowing whole blood. Digital image processing was used to analyse the real-time microscopic images obtained, resulting in measurements of morphological features and the number of platelets that compose each thrombus. The three-dimensional structures of the thrombi were also estimated. A photodiode was used to measure integrated end-point platelet accumulation at different axial positions along the surface. Additionally, the convection-diffusion equations were solved to estimate the concentrations of adenosine diphosphate, thromboxane A2, and thrombin generated by activated platelets at the blood-contacting surface.     

Platelets, inflammatory cells, von Willebrand factor, syndecan-1, fibrin, fibronectin, and bacteria co-localize in the liver thrombi of Bacillus anthracis-infected mice

Vascular dysfunction and thrombosis have been described in association with anthrax infection in humans and animals but the mechanisms of these dysfunctions, as well as the components involved in thrombi formation are poorly understood. Immunofluorescent microscopy was used to define the composition of thrombi in the liver of mice challenged with the Bacillus anthracis Sterne spores. Lethal infection with the toxigenic Sterne strain, in contrast to the non-lethal, non-toxigenic delta-Sterne strain, demonstrated time-dependent increase in the number of vegetative bacteria inside the liver sinusoids and central vein. Massive appearance of thrombi typically occluding the lumen of the vessels coincided with the sudden death of infected animals. Bacterial chains in the thrombi were stained positive for syndecan-1 (SDC-1), fibronectin, and were surrounded by fibrin polymers, GPIIb-positive platelets, von Willebrand Factor (vWF), CD45-positive leukocytes, and massive amount of shed SDC-1. Experiments with human umbilical vein endothelial cells (HUVECs) demonstrated the active role of the host response to the secreted pathogenic factors of bacteria during the onset of the pro-thrombotic condition. The bacterial culture supernatants, as well as the isolated proteins (the pore-forming toxin anthrolysin O and phospholipase C) induced release of vWF, while anthrolysin O, sphingomyelinase and edema toxin induced release of thrombin from HUVECs and polymerization of fibrin in the presence of human plasma. Conclusion: Our findings suggest that activation of endothelium in response to infection can contribute to the formation of occlusive thrombi consisting of aggregated bacteria, vWF, shed SDC-1, fibrin, activated platelets, fibronectin and leukocytes.     

Thrombogenesis of the Rabbit Arterial Plaque: An Electron Microscopic Study

Rabbit arteries, de-endothelialized with an intravascular balloon catheter and allowed to heal for 4 weeks, showed intimal changes that were similar to the preatherosclerotic fibromusculoelastic plaques of man. Reinjury of the healed vessels by balloon catheter produced marked quantitative and qualitative alterations of hemostasis, as compared to that in previously uninjured vessels. The most apparent modification of thrombogenesis 10 minutes after injury to the plaque was a large increase in the size of the thrombotic deposits. Features of this exaggerated response were the major participation of fibrin in thrombus formation and greater platelet accumulation. Some platelets and fibrin strands appeared to penetrate into and beneath the neointima. By 3 hours, these deposits had diminished in size, although the hemostatic mass remained larger in the doubly injured vessels.