| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1998;66:1191-1197
© 1998 The Society of Thoracic Surgeons
a Division of Thoracic and Cardiovascular Surgery , Mayo Clinic and Mayo Foundation, Rochester, Minnesota, USA
b Department of Physiology and Biophysics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota USA
c Department of Surgery Research, University of Arizona, Tucson, Arizona USA
Address reprint requests to Dr Schaff, Mayo Clinic, 200 First St SW, Rochester, MN 55905
Presented at the Poster Session of the Thirty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 26–28, 1998.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Autologous endothelial cells were sodded onto the luminal surface of the Permaflow conduit and immediately implanted to bypass the left circumflex coronary artery in adult mongrel dogs (n = 9). Unsodded Permaflow conduits were implanted as controls (n = 8). At 3 weeks, grafts were explanted and examined by scanning electron microscopy and immunostained for canine von Willebrand factor.
Results. Sodded grafts contained a confluent endothelial cell layer devoid of adherent thrombus or platelets and stained positively for canine von Willebrand factor. Unsodded grafts contained no endothelium and retained adherent platelets, collagen, and fibrin. Effluent from sodded grafts stimulated with calcium ionophore A23187 caused a significantly greater relaxation of its bioassay ring than effluent from unsodded grafts (60% ± 21% versus 12% ± 5%; n = 8, p < 0.03).
Conclusions. Sodding of endothelial cells onto a Permaflow coronary artery bypass graft results in a confluent, viable, nonthrombogenic, endothelial monolayer and releases vasodilator substances in response to calcium ionophore A23187. Endothelial sodding may optimize prosthetic grafts.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
The primary mechanisms of failure of synthetic coronary artery bypass conduits are anastomotic intimal thickening, low flow, and the inherent thrombogenicity of artificial polymers [1, 2]. The Permaflow coronary graft (Possis Medical Inc, Coon Rapids, MN) is a conduit designed to overcome the thrombogenicity of small diameter grafts by virtue of an arteriovenous fistula [3, 4]. Blood flow in the graft approaches 600 mL/min (Possis Medical Inc, unpublished technical data), and shear forces prevent platelet accumulation and discourage early thrombus formation.
Another approach to preventing thrombosis of synthetic grafts is endothelial cell transplantation to create a nonthrombogenic surface on the lumen of artificial polymers. First developed and reported by Herring and colleagues [5] in 1978, cell transplantation to vascular grafts accelerates formation of an antithrombotic lining on the blood flow surface.
Microvascular endothelial cell sodding is a single-stage technique wherein high-density falciform ligament (canine)- or adipose tissue (human)-derived microvessel endothelium is applied to the luminal surface of a prosthetic conduit immediately before implantation [6]. Little is known about endothelial sodding in grafts in the coronary circulation or in conduits with high blood flow and resultant shear forces, as occurs in the Permaflow graft [7]. This report describes successful sodding of microvascular-derived endothelium in a prosthetic conduit applied to the canine coronary circulation.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Under sterile conditions, a left thoracotomy was performed through the fourth intercostal space. A 3-cm segment of the left internal thoracic artery was isolated, followed by exposure of the proximal left subclavian artery. A pericardiotomy extended to the superior vena cava (SVC) and was controlled with umbilical tape. The proximal portion of the left circumflex coronary artery (LCX), beginning at its origin at the left main coronary artery, was exposed for a length of approximately 1.5 cm in the atrioventricular groove beneath the left atrial appendage. Heparin (2,000 U) was administered intravenously to maintain an activated clotting time of approximately 240 seconds, and the proximal left subclavian artery was controlled proximally and distally.
A 6-mm internal diameter, 30-µm internodal Permaflow graft was measured to the appropriate length. An arteriotomy was made in the subclavian artery, and the proximal portion of the graft was sutured end-to-side with running 7-0 polypropylene sutures. A rubber-shod clamp was placed on the graft 1 cm from the proximal anastomosis, and flow was reestablished in the left subclavian artery. Hemostasis was achieved, followed by the graft arteriotomy for the LCX anastomosis by using an elliptic hole punch (Scanlan Medical [St. Paul, MN] 2 x 4 mm). Heparin (3,000 U) was administered again to maintain an activated clotting time of approximately 350 seconds. Dogs then underwent coronary artery bypass grafting with the elliptic arteriotomy anastomosed side-to-side to the proximal LCX.
To avoid extracorporeal bypass, a temporary intraluminal coronary shunt maintained LCX myocardial blood flow during anastomosis of the prosthetic graft to the coronary artery in the beating heart, as described in detail elsewhere [8]. The proximal end of a 0.062-inch, inside diameter silicone catheter (Dow Corning Corp, Medical Materials, Midland, MI) was inserted through a small arteriotomy into the left internal thoracic artery, and pulsatile blood flow through the shunt was confirmed visually. Lidocaine (1 mg/kg) was administered intravenously, and the proximal LCX was ligated immediately beyond its origin from the left main coronary artery. An oblique arteriotomy was made in the LCX wall, with a distal suture loop around the LCX placed under tension to avoid back bleeding. The distal end of the silicone shunt was passed through the LCX arteriotomy to just beyond the looped suture but proximal to the origin of the first obtuse marginal branch of the LCX. Ties were secured and a third silk suture was positioned proximal to the tip of the shunt to prevent back bleeding when the LCX was opened. The shunt was unclamped and blood flow was reestablished into the distal LCX. Reestablishment of blood flow to the LCX region was confirmed visually by immediate resolution of myocardial cyanosis and resumption of regional myocardial systolic contraction. On average, coronary artery occlusion lasted less than 60 seconds.
A longitudinal arteriotomy was made in the LCX over the shunt, and the midportion of the Permaflow was anastomosed to the LCX in a bloodless field with running 7-0 polypropylene sutures. The clamp on the graft just distal to the subclavian anastomosis was removed, and air was flushed from the graft. A rubber-shod clamp was placed on the Permaflow distal to the coronary arteriotomy, the intracoronary shunt was removed, and the LCX was ligated proximal to the anastomosis. Blood flow into the distal LCX was completely dependent on the Permaflow graft. A partial occlusion clamp was applied to the SVC, and the distal end of the Permaflow graft was anastomosed to the SVC with running 7-0 polypropylene sutures (Fig 1 ). Blood flow was established through the graft and verified by palpation of a thrill distal to the Venturi valve. Also, blood flow through the coronary artery was documented by an ultrasonic flow probe (inside diameter 2 mm; Transonic Systems, Inc, Ithaca, NY), which was passed through the chest wall into a dorsal subcutaneous pouch. The pericardium was left open, and the thoracotomy was closed in layers. Each dog was allowed to recover for 21 ± 2 days, receiving intramuscular injections of torbugesic (1.0 mL) every 6 hours as needed for pain.
|
The vascular pellet was washed once by centrifugation with Dulbecco’s divalent, cation-free phosphate-buffered saline solution containing 0.1% bovine serum albumin. This final pellet was suspended in 9.5 mL of sodding media (media 199E without phenol red containing autologous canine serum; medium-to-serum ratio of 6:1). A fraction (0.5 mL) of the cell suspension was removed for Coulter counter analysis. The average number of cells isolated per gram of fat was 1.46 x 106. After resuspension of cells, each sodded graft received an average of 1.54 x 106 cells/cm2.
Graft preparation and sodding
Permaflow coronary artery bypass conduits in this study were supplied by Possis Medical Inc; these grafts are thin-walled, 6-mm inside diameter, and 30-µm internodal distance, with a distal silicone and titanium Venturi resistor. The average length used in each canine was 39 cm. Intraoperatively, each graft was measured to the appropriate length followed by placement of a metal cannula to the proximal end. The entire graft was then flushed with 20 mL of media 199 and 20 mL of sodding media using a glass syringe, followed by occlusion of the graft 2 cm distal from the anticipated site of the coronary anastomosis, creating an average sodded graft length of 13 cm. The graft and resistor distal to the rubber-shod clamp were not sodded and therefore served as an internal control.
Sodding medium (20 mL) was slowly injected (3 to 4 minutes) through the interstices of the proximal portion of the graft to be sodded. The cell suspension was then slowly introduced under minimal pressure while rotating the graft at 90° intervals to make three complete revolutions by the time cell injection was complete. Analysis by Coulter counter of the fluid that passed through the graft showed no cells, thus confirming sodding of the endothelial cells within the graft. Flushing continued with 20 mL of sodding medium being forced through the graft interstices at 2 to 3 psi and avoiding air introduction into the graft and sodding medium system. Grafts were placed without any manipulation as controls to reproduce current clinical practice. The total time necessary to perform cell isolation was 55 minutes, and graft sodding was completed in less than 7 minutes.
Graft explantation
After 21 ± 2 days the animals were returned to the operating room and anesthetized and the graft exposed through a median sternotomy. Coronary artery patency was documented by measurement of flow, and prosthetic graft patency was confirmed by intraluminal needle aspiration of fresh blood. The animals were killed by deep barbiturate anesthesia, and the graft, including the proximal and distal anastomoses, and heart were excised immediately. The grafts were washed gently with isotonic saline solution to remove loosely adherent cells.
One-centimeter segments of prosthetic conduit were obtained from the following regions: (1) 1.5 cm distal to the subclavian anastomosis; (2) 8 cm from the subclavian anastomosis, subsequently referred to as the "midgraft segment of the sodded conduit"; (3) 1.5 cm proximal to the coronary anastomosis; and (4) 8 cm proximal to the SVC anastomosis, termed the "venous portion of the sodded conduit." Each segment was prepared for light and scanning electron microscopy. Midgraft segments were also prepared for 
-smooth muscle cell actin or canine von Willebrand factor immunostaining.
Light microscopy
Samples of graft and surrounding tissue and native artery were fixed with 4% paraformaldehyde in phosphate-buffered saline solution, pH 7.4. Samples were dehydrated and fixed in paraffin, sectioned transversely in 5-µm segments, deparaffinized, and stained with hematoxylin and eosin and trichrome stains. Photomicrographs were obtained with a Nikon Optiphot light microscope.
Scanning electron microscopy
Graft segments for scanning electron microscopy were fixed with 3% glutaraldehyde in phosphate-buffered saline solution (pH 7.4), washed with phosphate-buffered saline solution, dehydrated in a graded series of acetone, and critical point dried. Dry samples were sputter coated by using a gold target. Samples were then examined and photomicrographs made with a JEOL 6400 scanning electron microscope.
Immunocytochemistry
Cryostat sections of explanted grafts were fixed on slides. Sections were reacted with primary antibodies to canine von Willebrand factor (provided by Dr James Catalfamo, New York State Health Department, Albany, NY) or -smooth muscle cell actin (Sigma Co, St. Louis, MO) and visualized by using peroxidase-conjugated secondary antibody. Nuclei were lightly counterstained with bromocresol green. Sections were then evaluated with a Nikon Optiphot microscope.
Superfusion bioassay
In 16 dogs (8 sodded and 8 unsodded), 4-cm segments of conduit from the midportion of the conduit between the subclavian and coronary anastomosis were evaluated for stimulated intraluminal release of endothelium-derived vasoactive substances. Grafts were excised and gently irrigated with a chilled, modified Krebs-Ringer’s bicarbonate solution.
Grafts were placed on a bioassay apparatus and care was taken to ensure flow through the graft was in the same direction in vitro as in vivo.
A deendothelialized coronary artery from an unmedicated dog (bioassay ring) was stretched to a resting tension of 10 g, found to be optimal in other experiments [10]. Responses of bioassay rings were examined during contractions stimulated by prostaglandin F2 (2 x 10-6 mol/L). The absence of endothelium on the bioassay rings was determined by absence of relaxation to acetylcholine (1 x 10-6 mol/L) or calcium ionophore (1 x 10-6 mol/L) during direct superfusion.
The bioassay rings were contracted with prostaglandin F2 and once the contraction was stable, an agonist was injected above the direct lines and the response was recorded. The ring was placed under the graft and agonist was injected to superfuse the graft and the bioassay ring. All grafts were stimulated with calcium ionophore A23187 (1 x 10-6 mol/L).
Data analysis
The results are expressed as mean ± standard error of the mean. Relaxations are expressed as percentage change in tension from the contraction of the bioassay ring to prostaglandin F2. All representative histology and immunocytochemistry data reported are from 1-cm proximal segments that were at least 1.5 cm from the subclavian anastomosis, 1-cm midgraft specimens (8 cm from the subclavian anastomosis), a 1-cm segment obtained 1.5 cm proximal to the coronary artery anastomosis, and a 1-cm segment obtained 8 cm distal to the coronary anastomosis. Immunocytochemistry data are reported from the midgraft segments. Statistical evaluation of the data was performed by Student’s t test for either paired or unpaired observations. Values were considered to be significantly different when the p value was less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Scanning electron microscopy
Ten sodded grafts were available for analysis. Eight grafts had a confluent cellular layer resembling endothelium devoid of adherent platelets, fibrin, or leukocytes (Fig 2A ). In the graft segment proximal to the Venturi tube, where the sodding procedure was not performed, all grafts revealed adherent thrombus and platelets with no evidence of a cellular layer (Fig 2B). One sodded graft contained no cellular layer, and platelets and thrombus were present on all segments of graft examined. One other sodded graft had incomplete sodding, with the absence of a cellular layer in the proximal segment, partial coverage of the midsegment, and complete coverage in the distal segment.
|
Light microscopy
Hematoxylin and eosin-stained cross-sections of sodded grafts revealed a confluent cellular monolayer (Fig 3 ). There were no confluent cells lining unsodded grafts.
|
-smooth muscle cell actin. In contrast, cells not in contact or beneath the luminal surface did not react with von Willebrand factor, but they reacted positively to -smooth muscle cell actin (Fig 4 ). Unsodded midgraft segments were devoid of immunostaining.
|
(1 x 10-6 mol/L) to the perfusate, 100 µL of calcium ionophore A23187 (1 x 10-6 mol/L) was infused through the stainless steel cannula; the absence of relaxation confirmed that there was no endothelium in the coronary artery bioassay rings. The perfusate from the stainless steel cannula (direct superfusion) was switched to the prosthetic graft (sodded or unsodded). In sodded conduits stimulated with 100 µL of calcium ionophore A23187 (1 x 10-6 mol/L), relaxation of the bioassay rings indicated intraluminal release of endothelial-dependent vasoactive compounds. The relaxation in bioassay rings perfused with effluent from sodded conduits stimulated with calcium ionophore was significantly greater (60.8% ± 21.2%; n = 8) than that found in unsodded conduits (12.1% ± 4.9%; n = 8; p < 0.03) (Fig 5 ).
|
Of eight unsodded conduits available for analysis, six were patent at all anastomoses. Two grafts had patent proximal and coronary anastomoses with an occluded distal anastomosis. There was no statistically significant difference in patency between sodded and unsodded grafts.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Relaxation of bioassay rings in response to stimulation of sodded conduits with calcium ionophore was significantly greater than relaxation observed from unsodded conduits, a finding that correlates with the morphologic and immunocytochemical evidence (reactivity with antibodies to canine von Willebrand factor) of a cellular layer resembling endothelium in sodded conduits. Our findings support previous reports characterizing endothelium from both sodding and seeding techniques applied to femoral and carotid artery grafts [11–13]. The demonstration that sodded cells release vasoactive factors suggests that these cells have the potential to maintain biologic activity that may inhibit thrombus formation, myointimal hyperplasia, and atherosclerotic degeneration.
Our data support the previous findings that autologous, falciform ligament, fat-derived endothelial cells can be obtained by using an operating room-compatible procedure [14] and subsequently can be implanted onto the luminal surface of vascular and coronary grafts in high densities. We deposited the cells onto the surface of the grafts by a pressure sodding technique, which takes advantage of the porous structure of expanded polytetrafluoroethylene grafts. This pressure sodding technique minimizes the time needed to deposit cells of the sodding inoculum onto the inner lining of the Permaflow conduit. By simply using the graft as a filter to trap cells in the sodding inoculum, we increased efficiency of cell deposition and greatly reduced the time required for cell implantation; gravity sodding adds an additional hour to the procedure [13], cell seeding techniques require 1 week for cell culture [11], whereas pressure sodding takes only 7 minutes.
Endothelial sodding differs from seeding technologies in several important respects. Seeding procedures use minimal cell numbers (owing to availability), which are placed within plasma or blood and subsequently preclotted within and on the graft surface. The resulting luminal surface exists as a thrombogenic fibrin meshwork; moreover, seeding covers a spectrum of techniques that generally involve macrovascular endothelial cell extraction in low densities (104 versus 106 cells/mL for sodding). Furthermore, cell seeding involves endothelial cell incubation and two separate procedures that compound patient inconvenience and potential risk as well as cost. Conversely, sodding uses a large number of cells to establish an endothelial cell monolayer on the luminal surface of a graft at the time of initial graft implant.
Most humans exhibit a rich supply of subcutaneous adipose tissue readily extracted through liposuction [15], but this is not usable in all of the animals previously studied. Thus, it is necessary to remove autologous fat using surgical excision [9]. Adipose tissue differs greatly in morphology, extent of microvascularity, and presence of nonendothelial cell types such as fibroblasts, mesothelium, and pericytes, depending on the source (eg, omentum or omental-associated fat, perirenal fat, or subcutaneous fat) [1, 9]. Canine falciform ligament adipose tissue is similar histologically to human subcutaneous fat, a source rich in endothelium once adipocytes have been removed. In our studies and those of others [9, 16], immunocytochemical analysis revealed that these cells express von Willebrand factor, suggesting endothelial origin.
To control for the possibility of spontaneous endothelialization as has been reported in canines [17], we used unsodded grafts that had no endothelium by histologic, immunocytochemical, and superfusion bioassay. Lack of endothelialization may have been the result of the high flow and shear forces regulated by the Venturi resistor [7]. Furthermore, the distal segments of sodded conduits (segments not treated by the sodding procedure) had no histologic evidence of an endothelium. The absence of endothelium on graft segments of sodded conduits not treated by the sodding procedure gives further credence to the efficacy of the sodding technique.
Previously, polytetrafluoroethylene grafts were used infrequently as coronary bypass conduits. We previously reported an expanded polytetrafluoroethylene circular, sequential coronary bypass conduit that was patent 12 years postoperatively; however, this appears to be a rare exception. Most clinical series report patencies as low as 14% at 45 months [18, 19].
The design of the Permaflow expanded polytetrafluoroethylene conduit may overcome some problems of small vessel conduits [4]. The Permaflow graft is used as an arteriovenous fistula, incorporating a Venturi resistor that controls flow to maintain proximal pressure [3, 7]. The coronary artery anastomoses are sewn side-to-side, and this configuration allows constant flow through the graft to prevent stasis and maintain patency, yet it allows systemic pressure for perfusion of the coronary arteries. The early clinical experience is favorable: 90% patency of coronary anastomoses at a mean follow-up of 13 months (28 patients and 73 coronary anastomoses). Phase II clinical studies are ongoing, and long-term study of graft patency must be completed before drawing conclusions on the efficacy of the Permaflow conduit [3, 5].
In summary, we have demonstrated the successful application of a confluent, nonthrombogenic, autologous endothelial surface to the lumen of a prosthetic coronary bypass conduit applied as an arteriovenous fistula, and report using the endothelial sodding procedure in the coronary circulation. Although early graft patency was not influenced, as demonstrated in this study, previous studies involving endothelial sodding have shown a statistically significant improvement in patency of grafts placed long-term (up to 1 year [9]) within the canine carotid and femoral arteries. The method of cell preparation and deposition described can be completed easily in the operating room and may be useful in clinical practice if longer term studies confirm the benefit of sodding prosthetic coronary expanded polytetrafluoroethylene grafts.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
