Development of an ex vivo model to investigate the effects of altered haemodynamics on human bypass grafts

The insertion of vein grafts into the arterial circulation may contribute to vessel wall thickening and accelerated atherosclerosis, a common feature of late vein graft failure. We aimed to develop a model suitable for investigation of the effects of altered haemodynamics on human saphenous vein following its implantation into the arterial circulation. Segments of human saphenous vein obtained from patients undergoing coronary artery bypass surgery were sutured at each end to PTFE and placed into a flow system. Pressure and flow rates to stimulate the arterial and venous systems were achieved. A theoretical model of the flow chamber was created and computational fluid dynamics software (FLOTRAN, Swanson Analysis Systems) was used to determine the flow profile within the model. In summary, a flow model has been developed to investigate the effect of altered haemodynamics on the molecular and pathological changes that occur in vein grafts incorporated into the arterial circulation.     

Development of an ex vivo model to investigate the effects of altered haemodynamics on human bypass grafts

The insertion of vein grafts into the arterial circulation may contribute to vessel wall thickening and accelerated atherosclerosis, a common feature of late vein graft failure. We aimed to develop a model suitable for investigation of the effects of altered haemodynamics on human saphenous vein following its implantation into the arterial circulation. Segments of human saphenous vein obtained from patients undergoing coronary artery bypass surgery were sutured at each end to PTFE and placed into a flowsystem. Pressure and flowrates to stimulatethe arterial and venous systems were achieved. A theoretical model of the flowchamber was created and computational fluid dynamics software (FLOTRAN, Swanson Analysis Systems) was used to determine the flowprofile within the model. In summary, a flow model has been developed to investigate the effect of altered haemodynamics on the molecular and pathological changes that occur in vein grafts incorpor-ated into the arterial circulation.     

Proliferative activity in stenotic human aortocoronary bypass grafts.

BACKGROUND: Aortocoronary bypass graft disease is responsible for long-term failure of autologous vein grafts. The analyses of proliferation and cell type characterisation in human bypass grafts harvested during re-do surgery make it possible to investigate the cellular processes leading to bypass graft failure. METHODS: 30 stenotic vein grafts and 25 control veins were explantated during re-do heart surgery procedures. The total area and cell count of the neointima, media and adventitia were calculated computer-assisted. Actively proliferating cells were identified using antibody to Ki-67 and positive cells were determined by double-label immunocytochemistry with SMC alpha-actin, CD 31 (endothelial cells), CD 68 (macrophages) and CD 45 (T-lymphocytes). RESULTS: Active proliferation was detected in different cell types with an average proliferation index of 0.15%, 0.18% and 0.086% for neointima, media and adventitia. Only 9% of proliferating cells in the neointima were SMC (not identified cells 40%); correspondingly, 14% SMC (not identified cells 33%) were detected in the media. Endothelial cells turned out to be the predominant proliferating cell type in all sections of the vessel wall. CONCLUSION: Proliferation in our series of stenotic vein grafts occurred at a low level, but was significantly higher compared to native control veins. While proliferation may play an important role in early lesions, our data clearly show low proliferation activity in advanced graft lesions. The identification of proliferating macrophages and T-lymphocytes implicate an additional inflammatory component in the development of human bypass graft disease. SUMMARY: To clarify the role of cellular proliferation in human aortocoronary bypass grafts, we characterized the cellular composition and proliferation index in 30 stenotic saphenous vein grafts in comparison to 25 native veins. Proliferation in our series of stenotic vein grafts occurred at a low level, but was significantly higher compared to native control veins.     

Impact of Shear Stress on Early Vein Graft Remodeling: A Biomechanical Analysis

In this study, an analytical technique was developed to estimate the dynamic, three-dimensional wall shear and tensile forces within the arterial circulation and applied to our established model of flow-mediated graft remodeling within the rabbit carotid system. Through selective unilateral distal ligation, vein grafts were exposed to distinct flow environments, characterized by a sixfold difference in mean flow rate. Implanted vein grafts were harvested at 1, 3, 7, 14, and 28 days to evaluate graft morphology. Hemodynamic and real-time imaging data, obtained at graft implantation and harvest, served as input for estimation of dynamic shear stress and wall tension. Marked differences in the remodeling process were observed in high vs. low flow grafts, with low flow grafts demonstrating accelerated intimal hyperplasia and reduced outward remodeling. The impact was a peak in the maximum and minimum shear stress at Day 7, with a delayed increase in lumen diameter leading to partial normalization of the wall shear by Day 28. Intramural wall tension reached a maximum at Day 3, with an increase in wall thickness leading to a significant reduction in these forces by Day 14. Despite the significant morphologic changes, no differences in the incremental moduli of elasticity were observed.     

Swirling flow pattern in a non-planar model of an interposition vein cuff anastomosis

One of the main causes of long-term failure of ePTFE grafts is the development of anastomotic intimal hyperplasia which leads to graft thrombosis. Experimental studies with bypass grafts have shown an inverse relationship between mean wall shear stress and intimal hyperplasia. The geometry of the anastomosis has a strong influence on the flow pattern and wall shear stress distribution. The aim of this in vitro study was to investigate the influence of non-planarity in a model of a distal anastomosis with interposition vein cuff, an anastomosis configuration that is increasingly being used because of improved clinical results. Laser Doppler anemometer measurements were carried out in silicone rubber models of interposition vein cuff anastomoses with planar and non-planar inflow. The pulsatile flow waveforms were typical of those found in femoro-infrapopliteal bypass. Axial and radial velocities were measured in the proximal and distal outflow segments. As expected a symmetrical helical flow pattern (Dean flow) was evident in the planar model. The model with non-planar inflow, however, gave rise to swirling flow in both the distal and proximal artery outflow segments for during the systolic phase. In patients, the anastomosis is usually non-planar. Since the configuration depends in part upon the tunnelling of the graft, this may be altered to some extent. Non-planar anastomotic configurations induce a swirling flow pattern, which may normalise wall shear stress, thereby potentially reducing intimal hyperplasia.     

Influences of graft diameter on the blood flow in 2-way bypassing surgery

The graft diameter plays a critically important role in the long-term patency rates of bypass surgery. To clarify the influence of graft diameter on the blood flows in the femoral 2-way bypass surgery, the physiologically pulsatile flows in two femoral bypass models were simulated with numerical methods. For the sake of comparison, the models were constructed with identical geometry parameters except the different diameters of grafts. Two models with small and large grafts were studied. The boundary conditions for the simulation of blood flow were constant for both models. The maximum Reynolds number was 832.8, and the Womersley number was 6.14. The emphases of results were on the analysis of flow fields in the vicinity of the distal anastomosis. The temporal-spatial distributions of velocity vectors, pressure drop between the proximal and distal toe, wall shear stresses, wall shear stress gradients and oscillating shear index were compared. The present study indicated that femoral artery bypassed with a large graft demonstrated disturbed axial flow and secondary flow at the distal anastomosis while the axial flow at its downstream of toe was featured with larger and more uniform longitudinal velocities. Meanwhile, the large model exhibits less refluences, relatively uniform wall shear stresses, lower pressure and smaller wall shear stress gradients, whereas it does not have any advantages in the distributions of secondary flow and the oscillating shear index. In general, the large model exhibits better and more uniform hemodynamic phenomena near the vessel wall and may be effective in preventing the initiation and development of postoperative intimal hyperplasia and restenosis.     

The use of external mesh reinforcement to reduce intimal hyperplasia and preserve the structure of human saphenous veins

The saphenous vein is the conduit of choice in bypass graft procedures. Haemodynamic factors play a major role in the development of intimal hyperplasia (IH), and subsequent bypass failure. To evaluate the potential protective effect of external reinforcement on such a failure, we developed an ex vivo model for the perfusion of segments of human saphenous veins under arterial shear stress. In veins submitted to pulsatile high pressure (mean pressure at 100 mmHg) for 3 or 7 days, the use of an external macroporous polyester mesh 1) prevented the dilatation of the vessel, 2) decreased the development of IH, 3) reduced the apoptosis of smooth muscle cells, and the subsequent fibrosis of the media layer, 4) prevented the remodelling of extracellular matrix through the up-regulation of matrix metalloproteinases (MMP-2, MMP-9) and plasminogen activator type I. The data show that, in an experimental ex vivo setting, an external scaffold decreases IH and maintains the integrity of veins exposed to arterial pressure, via increase in shear stress and decrease wall tension, that likely contribute to trigger selective molecular and cellular changes.     

Human Saphenous Vein Coronary Artery Bypass Graft Morphology,Geometry and Hemodynamics

Coronary artery bypass graft (CABG) failure has been linked to graft hemodynamics, in particular wall shear stress. This study characterizes the morphology, geometry and wall shear stress patterns in human CABGs. The intimal thickness (IT) in 49 human saphenous vein CABGs was measured by digital light microscopy. The geometry of six saphenous vein CABGs was replicated by post-mortem infusion of Batsons #17 anatomical corrosion casting compound. Graft hemodynamics were evaluated in two flow models, fabricated from the casts, under steady (Re = 110) and pulsatile flow (Re = 110, = 2) conditions. Saphenous vein CABGs in situ for more than 2 months had, on average, the greatest IT on the hood and suture sites of the distal anastomosis. Floor thickening was highly variable and significantly less than IT at the hood, suture site and graft body. All casts showed an indentation along the floor and 5/6 casts displayed a sharp local curvature on the hood. In both flow models, a large increase in wall shear rate occurred on the hood, just proximal to the toe. The local geometry of the hood created this large spatial gradient in wall shear stress which is a likely factor in hood intimal hyperplasia.     

Electron-microscopic detection of apoptotic and necrotic cell death in non-atherosclerotic areas of stenotic aortocoronary saphenous vein bypass grafts

Aortocoronary saphenous vein bypass grafts undergo structural alterations within the arterialized vein, resulting in graft stenosis and failure. Areas of the acellular intima contribute to fissuring, cracking and ulceration, while areas of the media become highly vascular but thinned. This study aimed to examine the ultrastructural features of cell death, including apoptosis and necrosis, in non-atherosclerotic areas of the stenotic aortocoronary saphenous vein bypass grafts. Thirteen stenotic vein grafts were obtained at redo coronary artery bypass grafting. The ultrastructural features of cell death were analysed by electron microscopy. Typical features of necrosis, including focal areas of cytoplasmic oedema, plasmalemmal destruction and nuclear condensation with cytoplasmic organelle destruction, were observed throughout the intima and media. Features of apoptosis, including the presence of apoptotic bodies, were also identified in the hyperplastic intima and its adjacent media. Our observations suggest that both apoptosis and necrosis occur in non-atherosclerotic areas of stenotic aortocoronary saphenous vein bypass grafts.     

Wall shear stresses in small and large two-way bypass grafts

Wall shear stress, as one of the most important hemodynamic parameters of the cardiovascular system, has been studied extensively in the numerical and experimental approaches to blood flow in various arteries. In order to clarify the influence of graft diameter on the wall shear stress in a femoral two-way bypass graft, the pulsatile blood flows in two models were simulated with the finite element method. Both models were constructed with different diameters of grafts. The main geometric structure and the boundary conditions were identical for both models. The emphasis was on the comparison analysis of wall shear stresses in the vicinity of the distal anastomosis. The temporal-spatial distributions of wall shear stresses, wall shear stress gradients, and oscillating shear index were analyzed and compared. The present study indicated that femoral artery bypassed with a large graft demonstrated relatively uniform wall shear stresses and small wall shear stress gradients, whereas it does not have advantages in the oscillating shear index. The large model exhibits better and more regular hemodynamic phenomena and may be effective in decreasing the probability of the initiation and development of postoperative intimal hyperplasia and restenosis. Thus, appropriately large grafts are applicable in the clinical practice of femoral two-way bypass operation. More detailed studies are necessary on this problem for the purpose of increasing the success rates of the femoral bypass grafts.     

Flow dynamics across end-to-end vascular bypass graft anastomoses

In this article, a numerical simulation of steady flow across an end-to-end vascular bypass graft anastomosis is presented.In vitro experiments were performed to determine the variations in the conduit cross section at the anastomosis. Penrose surgical drainage tubing was used to simulate an artery and was anastomosed with Polytetrafluoroethylene (PTFE) vascular grafts using a continuous suturing technique. Artery to artery anastomosis was simulated by suturing two Penrose tubing segments. The anastomotic specimens were subject to static transmural pressure in the physiologic range to determine the instantaneous diameter and compliance as a function of the distance from the anastomotic site. The experimentally determined geometries were used to simulate steady flow through an end-to-end anastomosis using the finite analytic (FA) numerical solution technique. The results demonstrated a region of flow separation 2 mm distal to the Penrose tubing-Penrose tubing anastomosis (simulating an artery-artery anastomosis) at higher transmural pressures. Moreover, wall shear stresses increased proximal to the anastomosis in flow from the Penrose tubing to the graft. In flow from the graft to the Penrose tubing, low wall shear stresses were observed distal to the anastomosis. Flow separation was observed distal to the anastomosis at higher transmural pressures with uniform inlet velocity condition. The region of low shear stress in flow from PTFE graft to the Penrose tubing was located nearer to the anastomosis with thin wall grafts than that with standard wall thickness grafts. Our steady flow model studies suggest a correlation between regions of low wall shear stress and the development of anastomotic neointimal fibrous hyperplasia in end-to-end anastomoses.     

Hemodynamic effects of the geometric dimensions of graft vessels in coronary artery bypass graft models.

The objectives of this investigation are to evaluate the rheologic properties in atherosclerotic disease treated with the various coronary artery bypass graft (CABG) models by numerical analysis, we used four different CABG models for the assessment of spatial fluctuation in wall shear stress, pressure variation and mass flow rate with Carreau model and Navier-Stokes equation. Wall shear stress was higher in a naturally tapered model (model 1) and a constant (non-tapered) diameter of the graft vessel the same as the distal LAD (model 4) than in others. Pressure variation along the native coronary artery and graft vessels was higher in a model 4, model 1 than in a reverse tapering model (model 2) and a constant diameter of the graft vessel the same as the proximal LAD (model 3). The mass flow rate of the distal part (kg/sec,.m(o)) was the highest in model 3. This study suggests that in vitro spatial simulation following CABG revealed that small caliber or tapered graft vessels have adverse hemodynamic effects on the native and graft vessels. By this technique it is possible to simulate the optimal distribution of local hemodynamic variables in patients treated with CABG, also to minimize the degeneration of graft vessel.     

管径比对全阻塞动脉旁路移植流场影响的数值研究

目的 研究管径比对全阻塞动脉旁路移植流场的影响,为指导动脉旁路移植手术,减少术后再狭窄提供理论依据。方法 采用数值方法研究5种不同移植管与主血管的管径比对全阻塞情况下动脉旁路移植流场的影响,分析速度、二次流、壁面切应力和壁面切应力梯度等血流动力学参数的分布及其随管径比增大的改变。同时,为表明本文所采用模型的合理性,针对目前常使用的两类模型,比较在管径比1.0情况下全阻塞完整模型（Model A）、全阻塞局部模型（Model B）和75％狭窄完整模型（Model C）之间的血流动力学差异。结果 Model A和Model C的血流动力学特性是完全不同的;移植管顶部截面内的速度分布对下游吻合处的主血管底部壁面切应力的影响是显著的,最大相差达79％。大管径比时,主血管底部的壁面低切应力区较大,但壁面切应力分布均匀,壁面切应力梯度较小。而小管径比时,主血管底部的壁面低切应力区较小,但壁面切应力梯度较大。结论 采用整体模型单独研究全阻塞情况下的管径比对流场的影响是有必要的。管径比对全阻塞动脉旁路移植的流场具有显著影响,采用大管径比进行动脉旁路移植将有助于缓解吻合口处由于再狭窄而产生的阻塞。     

Mouse vein graft hemodynamic manipulations to enhance experimental utility

Mouse models serve as a tool to study vein graft failure. However, in wild-type mice, there is limited intimal hyperplasia, hampering efforts to identify anti-intimal hyperplasia therapies. Furthermore, vein graft wall remodeling has not been well quantified in mice. We hypothesized that simple hemodynamic manipulations can reproducibly augment intimal hyperplasia and remodeling end points in mouse vein grafts, thereby enhancing their experimental utility. Mouse inferior vena cava-to-carotid interposition isografts were completed using an anastomotic cuff technique. Three flow restriction manipulations were executed by ligating outflow carotid branches, creating an outflow common carotid stenosis, and constructing a midgraft stenosis. Flowmetry and ultrasonography were used perioperatively and at day 28. All ligation strategies decreased the graft flow rate and wall shear stress. Morphometry showed that intimal thickness increased by 26% via carotid branch ligation and by 80% via common carotid stenosis. Despite similar mean flow rates and shear stresses among the three manipulations, the flow waveform amplitudes were lowest with common carotid stenosis. The disordered flow of the midgraft stenosis yielded poststenotic dilatation. The creation of an outflow common carotid stenosis generates clinically relevant (poor runoff) vein graft low wall shear stress and offers a technically flexible method for enhancing the intimal hyperplasia response. Midgraft stenosis exhibits poststenotic positive wall remodeling. These reproducible approaches offer novel strategies for increasing the utility of mouse vein graft models.     

A procedure to simulate coronary artery bypass graft surgery

In coronary artery bypass graft (CABG) surgery the involved tissues are overstretched, which may lead to intimal hyperplasia and graft failure. We propose a computational methodology for the simulation of traditional CABG surgery, and analyze the effect of two clinically relevant parameters on the artery and graft responses, i.e., incision length and insertion angle for a given graft diameter. The computational structural analyses are based on actual three-dimensional vessel dimensions of a human coronary artery and a human saphenous vein. The analyses consider the structure of the end-to-side anastomosis, the residual stresses and the typical anisotropic and nonlinear vessel behaviors. The coronary artery is modeled as a three-layer thick-walled tube. The finite element method is employed to predict deformation and stress distribution at various stages of CABG surgery. Small variations of the arterial incision have relatively big effects on the size of the arterial opening, which depends solely on the residual stress state. The incision length has a critical influence on the graft shape and the stress in the graft wall. Stresses at the heel region are higher than those at the toe region. The changes in the mechanical environment are severe along all transitions between the venous tissue and the host artery. Particular stress concentrations occur at the incision ends. The proposed computational methodology may be useful in designing a coronary anastomotic device for reducing surgical trauma. It may improve the quantitative knowledge of vessel diseases and serve as a tool for virtual planning of vascular surgery.     

动脉血流旋动原理在人造血管研制和血管移植术中的应用

目的 将血液流动的旋动流原理用于心血管介入治疗,以期解决小口径人造血管的急性血栓堵塞问题和搭桥手术后下游处血管内膜增生引起的血管再狭窄问题。方法 使用计算流体力学（CFD）分别研究具有旋动流特性的新型小口径、S型搭桥和偏心搭桥模型中的流场以及壁面剪切力的分布。同时研究在旋动流下血小板的黏附情况和不同角度的S型搭桥下的血管内膜增生。结果 旋动流能明显提高壁面剪切力,抑制血小板的黏附以及血管内膜增生。结论 在心血管介入治疗和器械设计中引入旋动流确实可明显改善这些器械中的血流流场,达到抑制小口径人造血管的急性血栓形成和搭桥手术后血管内膜增生的目的。     

Structural features of cell death in atherosclerotic lesions affecting long-term aortocoronary saphenous vein bypass grafts

Aortocoronary saphenous vein bypass grafts fail because of structural pathologies (thrombosis, intimal hyperplasia and atherosclerosis) within the 'arterialized' vein leading to graft stenosis. This study examined structural characteristics of atherosclerotic alterations in long-term aortocoronary artery saphenous vein bypass grafts with particular attention to the features of cell death in atherosclerotic lesions. Stenotic vein grafts were obtained from 10 patients at redo coronary artery bypass grafting operations. All the grafts were affected by histological abnormalities, with eight out of ten grafts showing evidence of atherosclerotic alterations in the intimal hyperplastic layer. Areas containing foam cells were examined by electron microscopy. Cells with cytoplasmic lipid accumulations were characterized by varying degrees of chromatin condensation, fragmentation or dispersion, by focal areas of oedema and vacuolisation of their cytoplasm, and by plasmalemmal destruction. Some lipid-filled cells exhibiting signs of destruction contained myofilaments and basal membrane fragments, allowing them to be identified as smooth muscle cells. Macrophage foam cells were found to have undergone similar destruction. No cells showing nuclear degeneration were observed to have intact cytoplasmic organelles. Neither were apoptotic bodies identified, but necrotic remnants were frequently seen. The results suggest that cell death in atherosclerotic lesions affecting aortocoronary artery saphenous vein bypass grafts occurs through oncosis rather than by apoptosis.     

Effect of a flow-streamlining implant at the distal anastomosis of a coronary artery bypass graft

Intimal thickening in the coronary artery bypass graft (CABG) distal anastomosis has been implicated as the major cause of restenosis and long-term graft failure. Several studies point to the interplay between nonuniform hemodynamics including disturbed flows and recirculation zones, wall shear stress, and long particle residence time as possible etiologies. The hemodynamic features of two anatomic models of saphenous-vein CABGs were studied and compared. One simulated an anastomosis with both diameter and compliance mismatch and a curvature at the connection, analogous to the geometry observed in a conventional cardiothoracic procedure. The other, simulated an anastomosis with a flow stabilizing anastomotic implant connector which improves current cardiothoracic procedures by eliminating the distal vein bulging and curvature. Physiologic flow conditions were imposed on both models and qualitative analysis of the flow was performed with dye injection and a digital camera. Quantitative analysis was performed with laser Doppler velocimetry. Results showed that the presence of the bulge at the veno-arterial junction, contributed to the formation of accentuated secondary structures (helices), which progress into the flow divider and significantly affect radial velocity components at the host vessel up to four diameters downstream of the junction. The model with the implant, achieved more hemodynamically efficient conditions on the host vessel with higher mean and maximum axial velocities and lower radial velocities than the conventional model. The presence of the sinus may also affect the magnitude and shape of the shear stress at locations where intimal thickening occurs. Thus, the presence of the implant creates a more streamlined environment with more primary and less secondary flow components which may then inhibit the development of intimal thickening, restenosis, and ultimate failure of the saphenous vein graft. © 2002 Biomedical Engineering Society. PAC2002: 8780Rb, 8719Rr, 8719Uv, 8763Lk, 8719Xx, 4262Be, 4780+v, 8710+e     

Mesh-constricted varicose and dilated veins used as arterial bypass grafts.

To avoid non-autogenous vascular prostheses in coronary and peripheral vascular procedures, otherwise unusable dilated, varicose or thin-walled veins can be implanted as naturally endothelialized grafts after being calibrated by insertion into mesh tubes. In six sheep the 14 mm diameter jugular vein was inserted into a 12 cm long, 6 mm diameter Dacron mesh tube and implanted as a femoropopliteal graft. On the contralateral side the femoral vein with a maximal diameter of 7 to 8 mm was implanted. At control angiography after two months all grafts were patent, none of the wrapped grafts showed wrinkles or signs of anastomotic hyperplasia. Segments of wrapped veins were casted in paraffin under physiological pressure and cut transversely. In all cases the vein wall was pressed against the mesh tubes without forming folds. Mesh-wrapped varicose veins were used for three femorocrural and seven femoropopliteal reconstructions. All grafts showed a smooth flow surface at control angiography. Indentations occurred only at sites where thick-walled side branches had been ligated rather than oversewn. One popliteo-pedal reconstruction occluded after two months and one patient with a crural reconstruction died two months after surgery with a patent graft. The remaining grafts were patent after 11 (2-32) months. Oversized veins can be considerably constricted by tubes without forming wrinkles. By this technique varicose veins can be used as bypass grafts and in general vein grafts can be matched in size to the recipient vessel.     

Surgical sealant in the prevention of early vein graft injury in an ex vivo model.

BACKGROUND: The amelioration of the adaptation process (arterialisation) of the vein graft wall to the arterial circulation in coronary artery bypass surgery by using extravascular support is clearly established in animal models and in in vitro and ex vivo set-ups. This support consists of some form of external graft-supporting modality like a prosthetic graft of stent. The clinical application of perivenous support, however, is hampered due to the fact that no easy applicable external support is available. Considering that application in the form of a spray is the most convenient modality, we evaluated whether polyethylene glycol is capable of providing adequate perivenous support. Polyethylene glycol is a synthetic, biodegradable product, used in cardiac surgery as a sealant, and is commercially available in the form of a spray. METHODS: Segments of human saphenous vein graft obtained during coronary artery bypass graft (CABG) procedures were placed in an ex vivo model, a side loop of the extracorporeal perfusion circuit, and perfused with autologous blood, making the circumstances identical to the implanted saphenous vein grafts concerning pressure, temperature, level of complement and leukocyte activation and blood pressure. Alternately around every other study vein graft segment polyethylene glycol was applied. Unsupported grafts served as control. After 1 min of solidification, perfusion was started with a pressure of about 60 mmHg (nonpulsatile flow). Perfusion was maintained for 60 min, after which the grafts were collected for light microscopy and electron microscopy. RESULTS: Light microscopy and electron microscopy showed remarkable attenuation of endothelial cell loss and less injury of smooth muscle cells of the circular and longitudinal layer of the media in the supported group compared to the nonsupported vein graft segments. CONCLUSION: Polyethylene glycol is able to provide adequate external vein graft support, preventing overdistension, in an ex vivo model. This provides a basis for clinical application. Further investigation is warranted to evaluate long-term effects.