Experimental vein grafts in the rat: re-endothelialization and permeability to albumin

Recombinant soluble apyrase APT102 inhibits thrombosis and intimal hyperplasia in vein grafts without adversely affecting hemostasis or re-endothelialization

Essentials New strategies are needed to inhibit thrombosis and intimal hyperplasia (IH) in vein grafts (VG). We studied effects of apyrase (APT102) on VGs and smooth muscle and endothelial cells (SMC/EC). APT102 inhibited thrombosis, SMC migration, and IH without impairing hemostasis or EC recovery. Apyrase APT102 is a single-drug approach to inhibit multiple processes that cause VG failure.

SUMMARY

Background Occlusion of vein grafts (VGs) after bypass surgery, owing to thrombosis and intimal hyperplasia (IH), is a major clinical problem. Apyrases are enzymes that scavenge extracellular ATP and ADP, and promote adenosine formation at sites of vascular injury, and hence have the potential to inhibit VG pathology. Objectives To examine the effects of recombinant soluble human apyrase, APT102, on platelets, smooth muscle cells (SMCs) and endothelial cells (ECs) in vitro, and on thrombosis and IH in murine VGs. Methods SMC and EC proliferation and migration were studied in vitro. Inferior vena cava segments from donor mice were grafted into carotid arteries of recipient mice. Results APT102 potently inhibited ADP-induced platelet aggregation and VG thrombosis, but it did not impair surgical hemostasis. APT102 did not directly inhibit SMC or EC proliferation, but significantly attenuated the effects of ATP on SMC and EC proliferation. APT102 significantly inhibited SMC migration, but did not inhibit EC migration, which may be mediated, at least in part, by inhibition of SMC, but not EC, migration by adenosine. At 4 weeks after surgery, there was significantly less IH in VGs of APT102-treated mice than in control VGs. APT102 significantly inhibited cell proliferation in VGs, but did not inhibit re-endothelialization. Conclusions Systemic administration of a recombinant human apyrase inhibits thrombosis and IH in VGs without increasing bleeding or compromising re-endothelialization. These results suggest that APT102 has the potential to become a novel, single-drug treatment strategy to prevent multiple pathologic processes that drive early adverse remodeling and occlusion of VGs.

Influence of the fiber diameter and surface roughness of electrospun vascular grafts on blood activation

Electrospun grafts have been widely investigated for vascular graft replacement due to their ease and compatibility with many natural and synthetic polymers. Here, the effect of the processing parameters on the scaffold's architecture and subsequent reactions of partially heparinized blood triggered by contacting these topographies were studied. Degrapol® (DP) and poly(lactic-co-glycolic acid) (PLGA) electrospun fibrous scaffolds were characterized with regard to fiber diameter, pore area and scaffold roughness. The study showed that electrospinning parameters greatly affect fiber diameter together with pore dimension and overall scaffold roughness. Coagulation cascade activation, early platelet adhesion and activation were analyzed after 2h of exposure of blood to the biomaterials. While no differences were found between DP and PLGA with similar topographies, the blood reactions were observed to be dependent on the fiber diameter and scaffold roughness. Scaffolds composed of thin fibers (diameter <1μm) triggered very low coagulation and almost no platelets adhered. On the other hand, scaffolds with a bigger fiber diameter (2-3μm) triggered higher thrombin formation and more platelets adhered. The highest platelet adhesion and activations rates as well as coagulation cascade activation were found in blood incubated in contact with the scaffolds produced with the biggest fiber diameter (5μm). These findings indicate that electrospun grafts with small fiber diameter (<1μm) could perform better with reduced early thrombogenicity due to lower platelet adhesion and lower activation of platelets and coagulation cascade.

Structural recovery of small arteries following clamp injury: a light and electron microscopic investigation in the rat

Endothelial injuries were induced in the left common iliac arteries (1 mm in diameter) of rats, by the placement of 1 mm Scovell-Lewis microvascular clamps for 5 minutes, to create a lesion in which to quantitate the rate and degree of cellular regeneration. The left (clamp-injured) and right (control) iliac arteries from the 15 rats used in this study were viewed with the electron microscope at 2, 7, and 14 days after clamping, and the clamp sites were analysed morphometrically. At 2 days there was only minimal denudation of the endothelium; most cells were disoriented and showed some signs of traumatic injury. By 7 days there was a completely continuous endothelial lining, but there was also evidence of increased cytoplasmic activity in these cells, as well as a statistically significant simplification in their intercellular junctional morphology. These changes persisted at 14 days after injury, but they were less pronounced. Smooth muscle cells in the media were relatively unaffected by the trauma in the first 2 days after clamping. However, they exhibited a change of phenotype from contractile to synthetic by 7 days after injury. By 14 days most smooth muscle cells had reverted back to the contractile phenotype, with little evidence of residual damage. These studies reveal that the reconstitution and regeneration of the endothelium is very rapid following clamp injury, but that significant residual ultrastructural changes in the interendothelial junctions persist for at least 14 days after injury. These findings indicate that there is potential for subsequent pathological changes in sites of vascular clamp injury.

Evans blue dye modifies the ultrastructure of normal and regenerating arterial endothelium in rats

Evans blue dye (EBD) identifies areas of increased vascular permeability, which is usually indicative of endothelial damage. Most studies examine EBD-stained areas light-microscopically, but others analyze the cells with the electron microscope. Electron microscopic studies have assumed that EBD itself did not change the ultrastructure of endothelial cells and this hypothesis was tested in the following study. The left iliac arteries of 20 rats were injured with 1-mm vascular clamps for 5 minutes. At 7 and 14 days after clamping, 10 rats for each time were infused intravenously either with normal-saline or EBD, perfused 30 minutes later with fixatives. Then the clamp-injured arteries, contralateral (unclamped) arteries, aortae, and the aortic bifurcations were removed for EM morphometry. In an additional (control) group of 10 rats, with no clamp injuries, 5 were infused with EBD and 5 with normal-saline and all 10 rats were perfused 30 minutes later, as above. EBD caused a significant simplification of the junctional morphology in both normal and regenerating endothelium. It also increased the area fractions of cytoplasmic vesicles in regenerating endothelium. These data demonstrate that EBD causes measurable ultrastructural changes in normal and regenerating endothelium. This effect should be taken into account when using EBD to assess various insults to blood vessels.

Mechanical stretching of human saphenous vein grafts induces expression and activation of matrix-degrading enzymes associated with vascular tissue injury and repair

After coronary artery bypass surgery, saphenous vein graft occlusion occurs through tissue remodeling. Although a likely trigger, the role of preparative mechanical injury incurred by the graft is not yet understood. We studied the early effects of simple mechanical injury on human saphenous vein grafts by exposing them to longitudinal stretch, a deformation which potentially occurs during surgery. We then maintained ex vivo for up to 7 days matched pairs of experimentally stretched and nonstretched (control) vein segments and examined the expression and activation of matrix metalloproteinases (MMPs) and integrin alphav, molecules implicated in vascular remodeling. At peak expression on day 3, stretched vein secreted 177 +/- 16% active MMP-2 (P < 0.01), 161 +/- 36% (P < 0.05) pro-MMP-9, and contained 206 +/- 18% (P < 0.01) alphav, a receptor for active MMP-2, compared to control. In situ gelatinase activity was present in the intima and adventitia of stretched veins, but not of control, and correlated spatially with expression of alphav. Stretch also increased severalfold cell proliferation (1.27 +/- 0.4 vs. 0.23 +/- 0.05% in control, P < 0.05), as assessed by bromodeoxyuridine incorporation. Furthermore, we found that cell proliferation colocalized with gelatinase activity and alphav in the adventitia. Our results show that a single longitudinal stretch of vein grafts produces significant changes in the expression and activation of key molecules in vascular remodeling. We also found support for the notion that the adventitial layer contributes to vein graft remodeling.

Vascular tissue adaptations in end-to-end autologous arterial grafts in rats: a morphometric analysis

Autologous vein grafts are employed extensively to bypass stenoses in the arterial circulation. More recently arterial segments have been used for such bypass surgery. In this study the adaptation of regenerating vascular tissues in experimental autologous artery grafts (4 mm long and 1 mm in diameter) in 20 adult male Wistar rats was analysed. At 1, 2, 4, 8 and 16 wk after insertion, 4 grafts per time interval were removed, processed for high resolution light microscopy and the thicknesses of the media and neointima, as well as the area fractions of smooth muscle cells, were analysed morphometrically. All grafts were reendothelialised by 2 wk. Neointimal hyperplasia (a subendothelial layer of smooth muscle cells) developed in all grafts and reached its maximal thickness (40.4 +/- 4.7 microns) at 2 wk. The area fraction of smooth muscle cells in the neointima of the artery grafts did not change significantly at any time from 2 to 16 wk. The media underlying the neointima of the artery grafts remained relatively constant throughout the 16 wk duration of the experiment. Whilst the total wall thickness of the grafts reduced significantly between 2 and 4 wk after insertion, at all times the grafts were thicker than the host artery.

Long-term structural alterations to endothelial cells in vein-to-artery grafts: a quantitative electron microscopic study

The intracellular structure of endothelium lining vein-to-artery grafts in rats was analysed, using transmission electron microscopy and morphometry, to determine the ultrastructural adaptations of endothelial cells in this altered vascular environment. Autogenous 4-mm sections of iliolumbar veins were inserted microsurgically into the left common iliac arteries of 16 male Wistar rats. At 3, 6, 26 and 52 weeks the cytoplasmic-vesicular, mitochondrial and rough endoplasmic reticular contents of endothelial cells lining the grafts, the opposite iliac arteries and the remaining ilio-lumbar veins were analysed morphometrically. There was a significant increase in the amount of all these cytoplasmic structures in endothelial cells at 3, 6 and 26 weeks, at 52 weeks there was also a significant increase in the volumes of mitochondria and cytoplasmic vesicles, but not in rough endoplasmic reticulum. It was concluded that the ultrastructure of endothelial cells lining these grafts is changed chronically after graft insertion, and we propose that this may be attributable to altered haemodynamic stresses within the graft.

Re-endothelialisation in autogenous vein grafts

OBJECTIVES

To clarify the course of re-endothelialisation (Re-E) in an entire graft and to establish the effect of immersion media for the preservation of endothelial cells.

METHODS

Autogenous femoral veins of dogs were immersed in heparinised saline solution (n = 18) or heparinized autogenous blood (n = 18). After immersion, the grafts were implanted into bilateral femoral artery, and were retrieved 1 day to 4-8 weeks after implantation.

RESULTS

For the grafts immersed in the heparinised saline solution, the values for % area of endothelial cell coverage before implantation, and at 1 day, 1 week, and 4 weeks after implantation were 44.9%, 6.2%, 14.5%, and 81.3%, respectively. For the grafts immersed in heparinised autogenous blood, the values were 73.5%, 20.6%, 79.2% and 95.5%, respectively. However, such relatively rapid speed of Re-E slowed down considerably after 1 week following implantation in this group.

CONCLUSIONS

The use of heparinized autogenous blood is strongly recommended as a preparation media for autogenous vein grafts. Almost all of the endothelial cells fall away in the earlier period after implantation and regenerate multifocally and irregularly. Re-E is incomplete even at 8 weeks after surgery, and we suggest that the area of incomplete Re-E may develop into intimal hyperplasia.

Late graft failure of autologous vein grafts for arterial occlusive disease: clinical and experimental studies

Late graft failure following arterial reconstructive surgery, especially after infrainguinal reconstruction, remains a major concern for vascular surgeons. To more effectively predict the outcome of reconstructed arteries, we herein propose an intraoperative flow waveform analysis which correlates well with the long-term patency rate of grafts. According to this flow waveform analysis, late graft failure was occasionally seen in grafts with type II waveforms when poor distal runoff vessels had been shown by the preoperative arteriogram. Next, to investigate which events occurring in autologous vein grafts under abnormal hemodynamics may contribute to late graft failure, a distal poor-runoff model was made in the canine femoral artery. In this review, we present the results of our investigation on autologous vein grafts using this poor-runoff model. We also relate our recent findings on the function of regenerated endothelium in autologous vein grafts.

The influence of nifedipine on microvascular vein graft intimal thickening

The influence of high-dose Nifedipine on the intimal thickening seen in experimental vein grafts was studied in an established rat model. The iliolumbar vein was grafted to the common iliac artery in 28 control and 30 Nifedipine-treated animals. There was one death in the control group and 10 deaths due to drug aspiration in the treated group; a further six control and four treated grafts were excluded from analysis due to occlusion or histological artefact. Animals were killed at 3 weeks, at which time a Nifedipine assay was performed. Longitudinal sections of the graft were prepared and the intimal thickness in the proximal, mid and distal graft measured using an eyepiece graticule. Median (range in brackets) intimal thicknesses for the control group were as follows: proximal 50 microns (10-120), mid 30 microns (10-70), and distal 30 microns (10-100). Results for the treated group were: proximal 65 microns (10-110), mid 30 microns (5-90), and distal 35 microns (20-120). There was no statistically significant difference between the two groups suggesting that Nifedipine is not of value in the inhibition of early vein graft intimal thickening.

Novel microsurgical model of experimental vascular neointimal hyperplasia

Neointimal hyperplasia is a common finding after trauma to blood vessels and also as a primary change in atherosclerosis. In this study we have developed a simple model, using microsurgical techniques, for the initiation of neointimal hyperplasia in the rat. In 24 Wistar rats, a 2 mm-diameter arteriotomy in the aorta was repaired with a "patch" of iliolumbar vein, using eight evenly spaced 10-0 Ethilon sutures. The patch overlapped the edges of the arteriotomy, and the sutures fastened the patch to the subjacent aorta. At 2, 6 and 12 weeks after surgery, the venous patch grafts and segments of the adjacent aortae of eight rats were removed. One-half of the specimens were analyzed by scanning electron microscopy and the other one-half by light microscopy. All patch grafts were surgically successful. Endothelial cells regenerated to cover the patch within 2 weeks of insertion. By 6 weeks after surgery, neointimal hyperplasia, consisting predominantly of smooth muscle, had developed in all patches to a thickness that was not significantly different from that of the adjacent aorta. These findings are consistent with data from other more complex experimental models of neointimal hyperplasia in the rat. We consider that this venous patch technique is a simple but effective model for the initiation of neointimal hyperplasia in the rat and may easily be used to study the experimental effects of various injurious or therapeutic agents on neointimal hyperplasia.

Vein to artery grafts: a morphological and histochemical study of the histogenesis of intimal hyperplasia

Failure of vein to artery grafts has been associated with intimal thickening (hyperplasia) and atherosclerosis. Current theories of intimal development, derived from arterial studies, show that smooth muscle cells migrate from the media to the intima after endothelial damage, where they proliferate and produce intimal hyperplasia. However, little is known of the histogenesis of these lesions in vein grafts. Experimental ilio-lumbar vein to iliac artery autografts were placed in 52 rats and analysed by light microscopy and histochemistry from 2 to 140 days after surgery. On day 2 the grafts and adjacent artery were severely damaged. Regeneration of damaged arterial tissue occurred by day 5, and thickening was already evident in the arterial intima. The intimal cells had histochemical characteristics of smooth muscle. By day 15, this hyperplastic intima was continuous across the anastomosis from the artery into the graft. After day 28 a wedge of densely packed cells was present in the vein graft intima for approximately 2 mm into the graft. By day 140, all the grafts were fully re-endothelialized. Intimal hyperplasia was present in all grafts and varied in thickness from 3 to 20 cells. Histochemical staining of these cells showed them to be of smooth muscle origin.

Influence of prazosin on experimental vein graft intimal thickening

The influence of prazosin on experimental vein graft intimal thickening was studied in a rat model. The iliolumbar vein was grafted into the common iliac artery in a prazosin-treated group (n = 17) and a control group (n = 23). Three weeks after grafting the animals were killed and the grafts harvested. Longitudinal sections were prepared and the intimal thickness measured in the proximal, mid and distal graft. Intimal thickness varied considerably in both groups. Median (range) intimal thicknesses for the control group were: proximal 50 (10-120) microns; mid 30 (10-70) microns; and distal graft 30 (10-100) microns. Results for the prazosin-treated group were: proximal 70 (10-160) microns; mid 25 (5-60) microns; and distal graft 25 (5-135) microns. There was no statistically significant difference in median intimal thicknesses between the control and treated groups.

Experimental evaluation of the length of microvenous grafts under normal tension

Four groups of microvenous grafts of 2, 4, 6, and 10 cm were used to determine whether length under normal tension affects patency rates in bridging arterial defects. Tension was evaluated by allowing the grafts to assume their normal orientation and length by filling them with blood, under physiological pressure, after the completion of the proximal anastomosis. The length was estimated through the normal range of motion such that the graft would not be too loose and/or tortuous following final anastomosis. The patency rates in the four length groups of this study ranged between 80% and 83%, with no statistical difference between groups. No statistically significant correlation was found between graft length and patency rate. It was concluded that the length of venous autografts does not affect patency.

Significance of the endothelial lining in prevention of intimal thickening of autogenous vein grafts in dogs

To better comprehend the role of the endothelial lining in occurrence and development of intimal thickening in arterially implanted autogenous vein grafts, two models of canine femoral arteries were prepared. In the Group I model, the implanted autogenous vein graft was kept under a normal flow condition for 2 to 4 weeks after implantation, then was exposed to an abnormal flow (poor run-off). In case of a 3- to 4-week normal flow, intimal thickening was practically nil. Scanning electron microscopic studies showed that this 3- to 4-week period corresponded to that of re-endothelialization of the autogenous vein grafts. Immunohistochemical studies of fibrinogen distribution in the autogenous vein graft wall in the Group II model revealed that the permeation of fibrinogen was enhanced in case of an abnormal flow condition for about 2 weeks after the implantation. We interpret these observations to mean that intimal thickening was induced by an accelerated permeability in the presence of abnormal flow conditions until full re-endothelialization after the implantation.

The rate of re-endothelialization correlates inversely with the degree of the following intimal thickening in vein grafts. Electron microscopic and immunohistochemical studies

The distribution of re-endothelialization and the development of intimal thickening were investigated electron microscopically and immunohistochemically using saphenous vein grafts implanted into the femoral artery in dogs. Animals were divided into two groups according to a difference of preparation of the venous grafts before implantation: group 1 without storage, and group 2 with storage for 1 h in 0.9% sodium chloride solution containing papaverine and heparin. In group 2 endothelial cells were almost totally denuded at 1 day, while endothelial islands were always left on the back of valves. In group 1 small islands of surviving endothelial cells were occasionally observed away from the valves. By 1 week, the re-endothelialization of group 1 and group 2 extended to 71 +/- 14.7% and 47 +/- 9% (mean +/- SD) of the total luminal surface area, respectively, and spread mainly from the valves but not from the adjacent artery. Immunohistochemical study using anti-dog fibrinogen IgG suggested that increased permeability of the graft wall continued for one month, in spite of the endothelial covering. Intimal thickening was most pronounced at 1 month, especially near the anastomosis, but was less in the middle of the grafts. These results indicate that intimal thickening of the venous grafts is a response to injury which intimately depends on the rate of re-endothelialization.

Vein-to-artery grafts: the long-term development of neo-intimal hyperplasia and its relationship to vasa vasorum and sympathetic innervation

A series of 14 vein-to-artery grafts, 1 mm in diameter and 5 mm long, were inserted microsurgically into iliac arteries of rats. They were analysed histologically 8-18 months later and compared with control iliac arteries in the same rats. Neo-intimal hyperplasia developed and was measured in all grafts but the values did not significantly exceed the equivalent intimal plus medial thicknesses of control arteries. Vasa vasorum developed and were quantitated as the number of vessels per mm2 of neo-intima, but also did not differ significantly from control values. The density of sympathetic innervation was quantitated using fluorescent catecholamines. There was an overall significant increase in the long-term graft innervation compared with control arteries. These results show that such small vein grafts adapt and function very effectively, in a manner remarkably similar to the artery they replace, for long periods of time.

Long-term histological changes in 1 millimeter polytetrafluoroethylene (Gore-Tex) prosthetic arterial grafts

Vein grafts are used extensively to repair blood vessels. However, when suitable vein segments are unavailable, alternative graft materials must be used. This study tested the suitability of 1 mm diameter polytetrafluoroethylene (PTFE) grafts for small-calibre arterial replacements. Grafts of 4 mm length were inserted microsurgically into the iliac arteries of 26 rats. The grafts were removed between 3 and 22 months postoperatively and examined with light microscopy and scanning and transmission electron microscopy (SEM:TEM). Anastomotic intimal hyperplasia was minimal, consisting of a few smooth muscle cells extending 100-200 microns onto the graft. The rest of the PTFE was covered with endothelium. In a few isolated areas, there was a subendothelial layer of smooth muscle, 2-10 cells thick. The short segments used in this study were highly successful, with an overall long-term patency rate of 80%. These grafts showed no evidence of excessive neo-intimal hyperplasia.

Intimal hyperplasia: the permeation of serum-derived substance into the arterial autovein graft under abnormal blood flow

The etiology of intimal hyperplasia in autogenous vein grafts used for arterial replacement was experimentally studied in dogs. We previously developed an experimental model, which mimicked a human extremity with poor run-off of peripheral arteries. This model characterized the abnormal flow by a weak fluctuation of wall shear stress at the site adjacent to the vessel wall. In this model we found that an autogenous vein implanted into the femoral artery under such abnormal flow conditions exhibited remarkable intimal thickening; 200-400 micron at 1 month and 300-500 micron at 6 months. A fluorescence microscopic study revealed that Evans blue-albumin complex entered the inner wall of vein grafts transplanted in abnormal flow conditions until about 2 weeks after implantation, after which further permeation was no longer observed. A similar observation was made in the fibrinogen distribution of subendothelial tissue. These results showed that abnormal flow conditions enhance the permeation of certain substances into immaturely repaired intimal tissues of autovein grafts. This phenomenon is thought to be responsible for the proliferation of smooth muscle cells, leading to hyperplasia of autovein grafts.

Vein to artery grafts: a study of re-innervation in relation to neo-intimal hyperplasia

This study was undertaken to determine whether there is an association between the sympathetic re-innervation and development of neo-intimal smooth muscle hyperplasia in vein to artery grafts. Iliolumbar vein to iliac artery grafts were placed in 21 rats by microsurgical techniques. Graft innervation and neo-intimal thickness were examined at five time intervals between 1 and 32 weeks after grafting. Nerve fibres were demonstrated microscopically by formaldehyde-induced fluorescence of catecholamines. The degree of innervation was quantitated by counting the nerve profiles and this was compared with the amount of neo-intimal hyperplasia. The distance between adventitial nerve profiles and neo-intima ('diffusion' distance) was measured to determine whether there was a trophic interaction between the two. These data were compared with similar measurements in control iliac arteries in the same animals. Although the development of both graft innervation and neo-intimal hyperplasia occurred coincidentally, no definite quantitative association between the two was established.

A review of the proliferative behaviour, morphology and phenotypes of vascular smooth muscle

This article reviews the proliferative, structural and synthetic behaviour of vascular smooth muscle cells under a variety of conditions. It shows how some experimental procedures produce dramatic increases in smooth muscle cell proliferation and, in many cases, subsequent cell migration to the intimal layer. Possible control mechanisms influencing changes in such activity are discussed. The morphology, histogenesis and differentiation of vascular smooth muscle is reviewed, with particular emphasis on the differentiation of such cells into contractile or synthetic phenotypes. The significance of the synthetic phenotype is discussed in relation to the synthesis of intra- and extracellular components.

Vein to artery grafts. An experimental study of reinnervation of the graft wall

Iliolumbar vein to iliac artery grafts were placed in 21 rats by microsurgical techniques. Graft innervation was examined at five time intervals between 1 and 32 weeks after surgery. Nerve fibers were demonstrated microscopically by formaldehyde-induced fluorescence of catecholamines. The morphology and degree of graft innervation were assessed, semiquantitatively, relative to the contralateral iliac artery (control) within each animal. Nerves were seen in the graft region as early as 2 weeks, but it was not until 4 weeks that they were present along its length (5 mm). The formation of a nerve plexus in the adventitia surrounding the graft was evident at 8 weeks. By 16 weeks the degree of innervation in the graft had increased to a level that was greater than the control iliac artery in three of four animals examined. Grafts at 32 weeks were also hyperinnervated. However, the morphology of this innervation was different from the control arteries; nerve fibers were finer, not varicosed, and were located at a greater distance from the outer layer of smooth muscle cells. The origin of the nerves appeared to be collateral sprouts from nerves supplying the adjacent iliac vein and also from invading vasa vasorum. The host iliac artery nerve plexus did not contribute to graft innervation.