| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2008;136:717-725
© 2008 The American Association for Thoracic Surgery
Evolving Technology |
a Christiaan Barnard Department of Cardiothoracic Surgery, University of Cape Town, Cape Town, South Africa
b Materials and Bioscience Center, Medtronic Inc, Minneapolis, Minn
c Medtronic Vascular, Danvers, Mass
Received for publication September 30, 2007; revisions received January 23, 2008; accepted for publication February 15, 2008. * Address for reprints: Peter Zilla, MD, PhD, Chris Barnard Department of Cardiothoracic Surgery, University of Cape Town, 7925 Observatory, South Africa. (Email: peter.zilla{at}uct.ac.za).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: The effect of mesh size on intimal hyperplasia and endothelial preservation was assessed in bilateral femoral interposition grafts in Chacma baboons (n
= 32/n = 8 per mesh size). No mesh support (group I) was compared with external nitinol meshes at three different sizes: loose fitting (group II), 25% diameter constricting (group III), and 50% diameter constricting (group IV). Mesh sizes were seen not only in isolation but also against the background of anastomotic size mismatch at implantation, expressed as quotient of cross-sectional area of host artery to vein graft (QC).
Results: Significant amounts of intimal hyperplasia were found in group I (QC median 0.20; intimal hyperplasia 6 weeks = 1.63 ± 0.34 mm2; intimal hyperplasia 12 weeks = 1.73 ± 0.5 mm2) and group II (QC median 0.25; intimal hyperplasia 6 weeks = 1.96 ± 1.64 mm2; intimal hyperplasia 12 weeks = 2.88 ± 1.69 mm2). In contrast, group III (QC median 0.45; intimal hyperplasia 6 weeks = 0.08 ± 0.13 mm2; intimal hyperplasia 12 weeks = 0.18 ± 0.32 mm2) and IV (QC median 1.16; intimal hyperplasia 6 weeks = 0.02 ± 0.03 mm2; intimal hyperplasia 12 weeks = 0.11 ± 0.10 mm2) showed dramatically suppressed intimal hyperplasia (P < .01) at both time points. Endothelial integrity was only preserved in group IV (P < .05). There were no significant differences in vascularization and inflammation in either interlayer or intergroup comparisons.
Conclusion: By using an animal model that addressed the clinical phenomenon of diameter discrepancy between vein graft and bypassed artery, we could demonstrate that suppression of intimal hyperplasia required constrictive mesh sizes.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The main pathologic processes behind failing vein grafts are intimal hyperplasia2 and atherosclerotic degeneration.3 Inasmuch as intimal proliferation typically precedes atherosclerotic degeneration4 it can also be seen as the "soil" for atherosclerotic vein graft changes.5 Therefore, subintimal tissue formation ultimately holds the key to the disappointing performance of vein grafts. A clinical consequence of the unsatisfactory patency rate of vein grafts was an increasing use of arterial grafts. Despite this trend, however, vein grafts remain the most widely used bypass conduits. Therefore, improving the patency of vein grafts is still a highly desirable goal. Understanding what causes the formation of vein graft intimal hyperplasia holds a key toward this goal.
Among the many factors contributing to the development of flow-limiting intimal hyperplasia, two are of central importance: low shear stress at the endothelial/blood interface6 and high circumferential wall stress.7 In a defined system, Pouiseille's law stipulates that circumferential stress increases and shear stress decreases with increasing diameter. Therefore, the relatively large diameter of clinically available vein grafts as compared with their target arteries–especially as seen in coronary surgery–is in itself a trigger for subintimal tissue proliferation.6,7 Additional dilatation of vein grafts after implantation further aggravates the situation. Vice versa, diameter reduction together with a decrease in wall tension can be expected to have an inhibitory effect on the formation of intimal hyperplasia.
As early as in 1963, Parsonnnet, Lari, and Shah8 proposed the concept of external mesh support with the goal of constricting the vein graft diameter to that of the artery, as well as preventing dilatation. Since then, the principle of vein graft protection with external meshes has been repeatedly revisited, but the multitude of animal models, materials, and diameters used resulted in opposing recommendations.9-11 To clarify a situation that has a significant potential to improve the clinical outcome of bypass surgery, we adopted a nonhuman primate model that mimicked the diameter mismatch of vein grafts and host arteries of patients. By comparing mesh sizes that created a spectrum, from loose fitting to a 50% diameter reduction, we attempted to determine an optimal size for vein grafts. Furthermore, by using the hyperelastic shape-memory material nitinol for the first time, we added a mechanical resilience factor that had not been addressed in the past.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Subsequently, non–mesh-supported controls were compared with 3 groups of meshes: loosely size matched (6.7 mm; n = 8), moderately constrictive (
25% diameter reducing; 5.0 mm; n = 8), and distinctly constrictive (50% diameter reducing; 3.3 mm; n = 8). Grafts were analyzed after 6 and 12 weeks of arterial implantation.
The incremental approach of first pairing nonsupported control grafts with 5.0-mm mesh-supported grafts in one set of animals followed by pairing of 3.3-mm meshes with 6.7-mm meshes in a subsequent set, followed by comparison across all 4 groups, was justified by 1-way analysis of variance (ANOVA) testing of in situ (preharvest) quotients of cross-sectional areas (Qc) of vein grafts to target arteries that confirmed a priori homogeneity of variance.
Vein Source Comparison
Chacma baboons do not have an equivalent to the human greater saphenous vein ("magna") but do have one corresponding with the lesser saphenous vein ("parva"). To determine the relevance of available baboon veins in the context of human saphenous vein grafts, histologic architecture and dimensions of human greater saphenous veins of 7 prospectively chosen patients were compared with those of the lesser saphenous veins (BSVs) and superficial femoral veins (BFVs) of 7 baboons (30.5 ± 4.3 kg) unrelated to the implant group. The effect of wall compression against an external mesh was assessed by pressure fixation (120 mm Hg) inside 3.3-mm nitinol meshes.
External Nitinol Meshes
Nitinol wires (BB Ni-Ti alloy: nickel 56.0 wt%, titanium 43.9365 wt%, carbon 0.033 wt%, oxygen 0.028 wt%, hydrogen 0.0025 wt%; thickness: 25/50 µm; Fort Wayne Metals, Fort Wayne, Ind) were braided using a 48-carrier braiding machine (Medtronic Vascular, Danvers, Mass [3.3 mm and 6.7 mm/50 µm] and a 72-carrier braiding machine (Prodesco Secant Medical, Perkasie, Pa [5.0 mm/25 µm]). Comparable wire density was achieved in all 3 diameters by choosing 24, 36, and 48 wires per circumference for the 3.3-mm, 5.0-mm, and 6.7-mm braid. The braid angle and pitch were 125° and 0.304 mm, respectively. Compliance was measured by circumferential tensile testing (Instron 5544 [Instron, Norwood, Mass], distilled water, 37°C). All 3 mesh diameters had a negligible compliance at 120/80 mm Hg of 1.14% ± 0.065%/100 mm Hg (3.3 mm), 2.06% ± 0.04%/100 mm Hg (5.0 mm), and 2.19% ± 0.12%/100 mm Hg (6.7 mm). The mesh diameters at 120/80 mm Hg filling pressures were calculated from the nonpressurized diameters (0 mm Hg) and the radial compliance of each mesh size, assuming a linear increase in diameter with an increase in pressure. The inner diameters of the meshes at 120/80 mm Hg were 3.39/3.37 mm (3.3-mm meshes), 5.12/5.08 mm (5.0-mm meshes), and 6.86/6.80 mm (6.7-mm meshes). Mesh dilatation at explantation was determined on the basis of cross-sectional image analysis of resin-embedded, pressure-fixed midgraft cross sections (Nikon E 1000 0.5x macro-lens/Leica QWin Pro V2.5; Nikon, Tokyo, Japan; Leica Microsystems, Wetzlar, Germany) pressure-fixed at 120 mm Hg.
Experimental Surgery
Adult chacma baboons (n = 16; 18.1 ± 0.9 kg) were provided by the breeding and quarantine facilities of the South African Medical Research Council. All experiments were approved by the Health Sciences Faculty Animal Ethics Committee of the University of Cape Town. Surgery and preoperative and postoperative care complied with the "Principles of Laboratory Animal Care" and the "Guide for the Care and Use of Laboratory Animals" (NIH publication No. 86-23). Under general anesthesia, reversed superficial femoral veins—taken between the adductor canal and the confluence with the deep femoral vein—were used as bilateral femorofemoral interposition grafts. In the mesh-supported groups, the veins had been gently pulled through an introduction tube inside the mesh before grafting.
Macroscopic Graft Dimensions
At implantation, measurements were taken with a caliper in the middle of the segment assigned for excision or after interposition grafting using a "no-touch" technique. Arteries and grafts were measured during systolic vessel expansion. The quotient of cross-sectional areas (QC = ah/ag whereby ah represents the cross-sectional area of the host vessel and ag that of the interposition graft) was chosen to express the dimensional mismatch between vein graft and host artery and indirectly as a reflection of flow velocity.
At explantation, grafts were perfusion-fixed (10% formalin/120 mm Hg) before they were resected en bloc. Midgraft sections underwent macroscopic image analysis for luminal dimensions (QWinPro V2.5; Leica Microsystems Imaging Solutions).
Light Microscopy
Two series of 25 consecutive sections (3 µm; nitinol wires removed) were cut for paraffin histology. The series were approximately 300 µm apart and used for light microscopy (modified Movat's, elastic Masson's trichrome, and picrosirius red for diascopic polarization microscopy of collagen/Nikon E1000M and a Nikon Coolscope) and immunofluorescence (CD31; FVIII:rAg; Actin/Desmin; CD 68; Antibody8740/MAC387 and Ki-67 with a nuclear Dapi counterstain/Nikon 90i). Non-dewired midgraft sections were infiltrated with Technovit T8100 resin (Heraeus-Kulzer, Wehrheim, Germany) and cut with an Isomet Precision Saw (Buehler, Dusseldorf, Germany) before being stained with hematoxylin and eosin.12
Electron Microscopy
After perfusion fixation in formalin, samples for electron microscopy were postfixed in 2% glutaraldehyde in phosphate-buffered saline solution. After processing, scanning electron microscopy images were digitally captured by Orion V5.20 software (Eli, Chartleroi, Belgium) before undergoing image analysis. Transmission electron microscopy sections were viewed with a Philips TEM420 (Carl Zeiss, Veldhoven, The Netherlands) and images captured on a SIS Megaview III CCD camera (Olympus Soft Imaging Solutions GmbH, Münster Germany).
Microscopic Assessment
For image analyses, composite images were assembled from digital single frames captured at 4x or 10x magnification (Nikon Eclipse90i and Nikon Coolscope) using EclipseNet (Laboratory Imaging, Prague, Czech Republic) software. Light/polarized light toggle analyses between combined modified Masson's trichrome and picrosirius stains were used for image analysis of wall thickness and dimensions of wall layers. Intimal hyperplasia tissue was discernible both by its demarcation through a clear though sometimes fragmented internal elastic lamella and its distinctly weaker staining for actin and desmin than the adjacent media. The adventitial layer was defined between the outer demarcation of the media and the inner demarcation of the dense circumferential collagen structures characterizing the fibrous capsule. Precise mesh positions were determined on resin sections. Assessment of media muscularity, vasa vasorum, nuclear density, inflammatory cell infiltration, and cell proliferation, as well as verification of surface endothelialization, was based on immunofluorescence composite pictures. Entire wall layers were analyzed. Cellularity was related to the cross-sectional area of wall layers. Muscularity was defined as the partial cross-sectional area of smooth muscle cells within the boundaries of the media. Cellular phenomena such as cell proliferation (Ki67) and macrophage infiltration were related to the cellularity (Dapi positive nuclei per area) of each layer. Vasa vasora (number of vessels; partial cross-sectional area) were related to cross-sectional areas of wall layers (Figure 1
). For the assessment of surface endothelialization, the presence of endothelial cells on the blood surface was first immunohistochemically confirmed (CD31) before image analysis (Scion Image software; National Institutes of Health, Bethesda Md) of scanning electron micrographs was used for quantification.
|
For assessment of the differences between vein types, 1-way ANOVA of layer dimensions was followed by Tukey-Kramer honestly significant difference post hoc analysis.
Analysis of change in cross-sectional luminal areas excluded preimplant values from inferential analyses of the remaining time points since these referred to unrestricted in situ veins.
Comparison of layer dimensions before and after external mesh support was performed by the Student t test with 1-tailed P values being reported. Nonparametric testing using the Wilcoxon rank sum test was used for comparison of QC data between groups. ANOVA testing was used for the analysis of the 3 time points (implant, 6 weeks, and 12 weeks) with the Dunnet t test being used for post hoc analysis of values at 6 and 12 weeks versus implant control values. Similarly, cross-sectional anatomic layer areas were examined by ANOVA testing with post hoc between-group comparisons using the Student t test with Bonferonni correction of 2-tailed P values for multiple comparisons. Square root transformation of intimal hyperplasia area data was performed following the Shapiro–Wilk testing for normality. Patency of grafts was examined using Fisher's exact test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The histologic comparison of vessel wall layers showed the media of both baboon veins to be of similar composition and only mildly thinner than that of the human greater saphenous vein (23% BSV; P = not significant; 24% BFV; P = not significant). Furthermore, muscular contraction seemed to enable smooth caliber adjustment of the veins. None of the veins inside nitinol meshes of half the maximal vein graft diameter showed signs of folding or buckling.
Effect of External Mesh Support on Patency, Dimension, and Cross-sectional Quotient of Vein Grafts
Graft patency was 87.5% at 6 weeks and 75.0% at 12 weeks with no significant difference between the time points and/or the groups.
On exposure to the arterial circulation, vein grafts of the unsupported control group expanded by 24% outer diameter (OD) (P = .09). After a similar distention within the 6.7 mm meshes, the OD of vein grafts and the inner diameter (ID) of meshes were almost size matched (ODVein graft 3.6% ± 13.3% < IDMesh). In contrast, the 5.0-mm and 3.3-mm meshes led to a diameter reduction of vein grafts by 25.9% ± 12.3% and 48.7% ± 6.6%, respectively. As a result, the QC differed significantly (P < .0001; Wilcoxon rank sum test) between the control group (QC median = 0.20; range 0.18–0.42) and the 6.7-mm meshes (QC median = 0.25; range 0.23–0.47) on the one hand and 5.0-mm (QC median = 0.45; range 0.43–0.47) and 3.3-mm meshes (QC median = 1.16; range 1.13–1.18) on the other. After 6 weeks of implantation, control veins experienced a 63% cross-sectional narrowing of the lumen (P < .05) (Figures 2 and 3
), followed by a distinct 188% dilatation at 12 weeks (P = .027; Dunnet t test) (Figures 2 and 3). Although the meshes prevented such an extreme vein graft dilatation, some degree of mesh dilatation was also observed (Figure 3).
|
|
). As much as intimal proliferation was pronounced (Figure 6), it only accounted for17.4% ± 6.4% (unsupported) and 14.3% ± 7.9% (6.7-mm mesh supported) of the overall increase in wall thickness. The bulk of tissue formation took place in the fibrous capsule that "filled" the more than 0.5-mm wide space between the mesh and the adventitia. In contrast, a relatively thin vessel wall was regularly found in the 2 mesh-supported groups that constricted the vein grafts. At 12 weeks, cross-sectional appearances remained unchanged, but histologically there was a general decrease in medial and adventitia thickness.
|
|
|
Intima
Two typical pictures prevailed throughout: a distinct, irregular, and often half moon–shaped intimal tissue that was rich in extracellular matrix in both the control group and the 6.7 mm meshes as opposed to a near complete absence of intimal tissue formation in the constrictive 3.3 mm and 5.0 mm groups. An internal elastic membrane was regularly present (Figure 5). The most dramatic increase in intimal tissue occurred in the 6.7 mm mesh-supported group, where it continued beyond the initial 6 weeks of implantation (Figure 6). In nonsupported controls, intimal hyperplasia appeared to reach a cross-sectional equilibrium at 6 weeks, leading to a relative thinning of the neointima at 12 weeks, corresponding with graft dilatation.
Media
The doubling of medial thickness in controls and the 6.7-mm mesh-supported group during the initial 6 weeks was associated with an actual drop in muscularity from 60.9% ± 13.0% (at implant) to 33.3% ± 3.3% (controls; P = .0046) and 35.9% ± 3.8% (6.7-mm meshes; P = .0226). Overall, a diminished total amount of smooth muscle tissue was found within the media (–37.8%, P = .0015 in controls; –41.9%, P = not significant in 6.7-mm meshes; –81.2%, P < .0001 in 5.0-mm meshes, and –93.6%, P < .0001 in 3.3-mm meshes). Intimal cells showed signs of high secretory activity on transmission electron microscopy. Remnants of platelets were found within the extracellular matrix.
Adventitia, Fibrous Capsule, and Periadventitia
The two most striking features of these outer layers were the almost absent inflammatory response against the nitinol struts (except for a few foreign body giant cells) and the surprising abundance of elastin in the mesh-supported groups (between the mesh and the media as well as relatively far outside the mesh in the periadventitial layer). The fibrous capsule was distinctly thicker in 6.7-mm meshes than in other vein grafts after 6 weeks (280% vs controls, P = .0052; 198% vs 5.0-mm meshes, P = .0308; and 270% vs 3.3-mm meshes, P = .0089), while no further change in capsule thickness was observed in any of the groups thereafter.
Inflammatory Response
The 6-week macrophage index of the intimal tissue was higher in the 6.7-mm group than in any other group but followed the overall trend of an abating macrophage response at 12 weeks. No significant differences were found between the groups.
Cell Proliferation
The ratio of proliferating to resting surface endothelial cells was below 1.2% in all groups except in the 6.7-mm mesh-supported group, where it was 14.4% (6 weeks) and 2.4% (12 weeks) (no significant difference). Within the wall layers, the proliferative activity did not exceed 3% at any time point. The only significantly higher proliferation index in interlayer and intergroup comparisons was found in the fibrous capsule of the 6.7-mm mesh-supported group at 6 weeks (P = .0018).
Vascularization
In the 3.3-mm and 5.0-mm groups, the ingrowth of vasa vasorum was confined to layers outside the media, whereas the controls and the 6.7-mm groups showed scanty vessels also within the media and neointima, at both 6 and 12 weeks. Inasmuch as vessel density and cross-sectional area increased over time, this trend, as well as intergroup and interlayer comparisons, were not significant.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The main underlying paradigm of the study was that the basic biomechanical principles that regulate intimal hyperplasia in general also determine the effect of external meshes. Given the fact that low shear stress and high circumferential wall stress are prime switches for the development of intimal hyperplasia6 Pouiseille's law implies that the diameter mismatch usually encountered between a large vein graft and a small host artery is a main trigger for the development of intimal hyperplasia. Inversely, it also implies that constrictive mesh sizes should counter this trigger and thus exert a suppressive effect on intimal hyperplasia.
By using the relatively small saphenous veins for interposition grafts in large arteries such as the carotids, some of the previous studies created experimental flow conditions that were distinctly different from the clinical situation. Therefore, conclusions regarding the effectiveness of loose-fitting meshes11, 13-15 are not unconditionally applicable to human bypass surgery.
In coronary artery bypass surgery, for instance, reversed saphenous veins are most often used for grafting of the marginal branch of the circumflex artery and the posterior descending artery.
Given the average dimension of these coronary arteries (midsection ID/cross-sectional area 2.1 mm/3.4 mm2 and 1.8 mm/2.5 mm2, respectively16), as well as that of saphenous vein grafts (OD 4.18 ± 0.4 mm; wall thickness 458 ± 17 µm; cross-sectional area 8.4 mm2; 200 consecutive patients; unpublished data), the quotients of cross-sectional areas (QC) would be in the region of 0.30 to 0.40. Under clinical circumstances, where the runoff may be compromised, the QC may even be below 0.30. In contrast, if an internal thoracic artery (average OD: 3.01 ± 0.51 mm; wall thickness 248 ± 108 µm; cross-sectional area 6.0 ± 0.46 mm2; unpublished data) was used to graft the midsection of the left anterior descending artery (ID 2.6 mm; cross-sectional area 5.3 mm2
16), the cross-sectional quotient would be 0.88.
In our present study, the QCs for control grafts and size-matched mesh-supported grafts were therefore similar to the QCs estimated for clinical vein grafts. Hence, it was not surprising that both the control and size-matched groups showed a degree of diffuse early intimal hyperplasia typically found in the majority of coronary bypass grafts of patients.17 In contrast, when the QC was increased to more than 0.45 through diameter reduction of the vein graft, intimal hyperplasia was almost completely suppressed. This interdependent consideration of graft constriction and pre-existing diameter mismatch between vein graft and host artery may resolve the puzzle as to why some studies reported distinctly better results with constricting meshes9,10,18-21 while others saw a positive effect with lose-fitting meshes.11,13-15 Typically, those promoting constrictive mesh sizes used relatively large venous conduits with an estimated QC of between 0.15 and 0.4,9,10,18-21 whereas those seeing a benefit in oversizing used smaller vein grafts with an estimated QC of up to 0.7.11,13-15 It is therefore evident that those studies advocating the oversizing of external meshes used animal models in which the vein grafts were relatively narrow when related to the runoff vessels and thus experienced higher a priori shear forces at the endothelium–blood interface. It has been shown that the threshold value for local levels of wall shear stress is 5 dynes/cm2 below which intimal thickening occurs.22 With a QC of as high as 0.7, this threshold is not even remotely undercut, even if grafts remain unsupported. Therefore, constrictive mesh sizes would not have added an additional shear-stress benefit under these circumstances. The fact that differences were still seen between mesh-supported and unsupported grafts despite the high baseline QC may be explained with a "tamponading" effect inside the mainly used Dacron grafts caused by the known seroma buildup between graft and vein.23 Thereby, even oversized meshes would turn into expansion-limiting devices. The resulting lower tangential wall stress would show an effect even if the shear forces remained subthreshold throughout.
Alternatively, different degrees of wall vascularization and/or inflammation were suggested as explanations for the protective effect of external vein support. A high vasa vasorum density was seen as a key mechanism through which loose-fitting Dacron meshes suppressed the development of intimal hyperplasia. In return, the ineffectiveness of tight-sitting Dacron meshes was explained by a lack of vasa vasora.15 The near absence of a microvessel plexus inside such meshes was ascribed to insufficient space for the development of a neoadventitia between mesh and vein graft.23 Since the delicate and wide-open nitinol meshes, which we used for external support, allowed for a well-developed vasa vasorum plexus in the neoadventitia of all groups, the previously reported effect of constrictive meshes on vasa vasorum formation may well be specific for polymeric fabrics. The fact that we observed a near-complete suppression of intimal hyperplasia suggests that vasa vasora cannot be seen as exclusive mechanisms behind the effect of external vein meshes. Similarly, the level of inflammation seen in our nitinol mesh–supported grafts was distinctly lower than that described for synthetic24 or resorbable meshes of previous studies.11 The distinct inflammatory response to Dacron was offered as one explanation for the failure of tight-sitting meshes.
However, as much as the suppression of intimal tissue proliferation is crucial in vein grafts, one must not overlook the importance of endothelial preservation. In our present study, endothelial preservation was even more affected by mesh size than intimal hyperplasia. While a 25% diameter reduction was sufficient to largely suppress intimal hyperplasia, this moderate diameter reduction led to a similar degree of early endothelial cell detachment as in control veins and the size-matched mesh-supported group. In contrast, 50% diameter constrictions to QCs above 1.10 appeared to completely preserve the endothelium. This apparent correlation of a high QC with endothelial preservation seems to be in accordance with most previous studies: the most pronounced difference in endothelial integrity between mesh-supported grafts and unsupported controls was found in those studies in which the baseline QC seemed low9,20,25 and mesh support resulted in a significant constriction of the vein grafts. In contrast, when the baseline QC was already high, no significant endothelial loss was observed whether vein grafts were mesh supported or unsupported.14 This confirms a previously suggested link between low shear stress and endothelial detachment.26,27 It could explain why, in the present study, all 3 groups with shear forces below or equal to 5 dynes/cm2 showed substantial endothelial loss as opposed to the near complete endothelial preservation in the 50% diameter–constricted group, where the estimated wall shear stress was higher than 20 dynes/cm2.
Apart from intimal hyperplasia and endothelial preservation, another observation of the present study was a diminishing overall quantity of smooth muscle cells in the media. This atrophy of vascular muscle tissue puts a process often described as "arterialization" of vein grafts into perspective. Hence, increases in medial thickness may have been merely a reflection of a combination of bulk compression resulting from adventitial shrinkage and pronounced extracellular matrix deposition. Given the fact that veins reach their adventitial "arrest" point at as low a pressure as 30 mm Hg, loss of pulsatility may be one reason for this medial atrophy. This explanation would be supported by the even more dramatic net reduction in medial smooth muscle cells found in constrictive meshes where the noncompliant mesh absorbed the wall strain even more than the adventitial collagen in unsupported grafts.
In summary, we could demonstrate in a nonhuman primate model that moderately constrictive external vein meshes made of the hyperelastic shape-memory material nitinol are capable of almost completely abolishing the formation of intimal hyperplasia. If the goal of vein graft protection is also to preserve endothelial integrity, an even higher degree of diameter reduction seems to be required that mimics the cross-sectional area quotient between internal thoracic artery grafts and the left anterior descending coronary artery.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Re MS 07-1358, disclosure is needed, as follows: This study was supported by a grant from Medtronic; Peter Zilla and Michael Wolf hold patents on the technology described in this study; Nasser Rafie reports consulting fees from Medronic; Michael Wolf is an employee of Medtronic.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Schoettler, J. Jussli-Melchers, C. Grothusen, L. Stracke, F. Schoeneich, S. Stohn, G. Hoffmann, and J. Cremer Highly flexible nitinol mesh to encase aortocoronary saphenous vein grafts: first clinical experiences and angiographic results nine months postoperatively Interact CardioVasc Thorac Surg, October 1, 2011; 13(4): 396 - 400. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. T. Inderbitzin, P. Matt, F. S. Eckstein, and O. Reuthebuch eComment: External nitinol meshing of venous coronary artery bypass grafts: is safety of application really in doubt? Interact CardioVasc Thorac Surg, October 1, 2011; 13(4): 400 - 400. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Desai, J. Mirzay-Razzaz, D. von Delft, S. Sarkar, G. Hamilton, and A. M. Seifalian Inhibition of neointimal formation and hyperplasia in vein grafts by external stent/sheath Vascular Medicine, August 1, 2010; 15(4): 287 - 297. [Abstract] [PDF]
|
|
|
|
|
|
G. Rescigno, A. Angelini, A. D'Alfonso, and L. Torracca Coronary graft use of a new external mesh support Interact CardioVasc Thorac Surg, April 1, 2010; 10(4): 645 - 647. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Human, T. Franz, J. Scherman, L. Moodley, and P. Zilla Dimensional analysis of human saphenous vein grafts: Implications for external mesh support. J. Thorac. Cardiovasc. Surg., May 1, 2009; 137(5): 1101 - 1108. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
) and loose collagen-rich media in controls and the size-matched group. Distinct (50%) downsizing abolished intimal hyperplasia tissue (IH) and led to a thin, compacted media.