- Email: [email protected]
Qualitative microscopy of implanted vein grafts
Qualitative microscopy of implanted vein grafts Effects of graft integrity on morphologic fate This is an investigation of the relationship between graft preparation techniques and the subsequent fate of vein grafts. Vein grafts intentionally injured by warm saline storage demonstrated endothelial and smooth muscle cell damage. In the acute postimplantation period, platelet adhesion/activation and white cell infiltration were present. By 7 days the endothelium had "healed," but the underlying smooth muscle cells had modulated and were of the synthetic phenotype. This persisted at 30 days, but by 60 days the graft wall remodeled with smooth muscle cells that were of the contractile phenotype, with an organized extracellular matrix. None of these injurious responses were noted in optimally prepared papaverine-treated vein grafts. Optimal preparation of vein grafts is effective in minimizing endothelial and smooth muscle cell injury before and after arterial reconstruction. Prevention of vein graft injury during harvesting prevents the morphologic changes characteristic of the "arterialization response." (J THORAC CARDIOVASC SURG 1992;103:671-7)
William C. Quist, MD, PhD, Christian C. Haudenschild, MD, and Frank W. LoGerfo, MD, Boston, Mass.
h e saphenous vein continues to be the optimal graft for infrapopliteal and complete cardiac revascularizations.'- 2 Despite this, a significant number of vein graft failures occur in the acute and delayed postoperative period. Endothelial cell injury and smooth muscle cell proliferation appear to be important factors leading to these vein graft failures. Early studies from this laboratory implicated harvesting techniques as a mechanism for both endothelial and smooth muscle cell damage before arterial reconstruction. Postulating a relationship between graft integrity at the time of implantation and subsequent graft function, we developed a technique that led to optimal preservation of both endothelial and smooth muscle cells before arterial reconstruction.' Graft integrity was best preserved with transcutaneous perivenous papaverine, controlled temperature and pressure during distention and storage, and maintenance of distention during the storage and reconstructive pericds.l 4 From the Divisionof Vascular Surgery, Harvard Medical School, New England Deaconess Hospital, and the Mallory Institute of Pathology, Boston University School of Medicine, Boston, Mass. Received for publication Aug. 22, 1990. Accepted for publication Feb. 7, 1991. Address for reprints: Frank W. LoGerfo, MD, New England Deaconess Hospital, 110 Francis St., Boston, MA 02215.
12/1/29023
The purpose of this study was to use qualitative microscopy to compare serially the effects of vein graft preparation techniques on the subsequent morphologic events after arterial reconstruction. If vein graft preparation technique has an impact on morphologic events after arterial reconstruction, a significant number of vein graft failures may be prevented and the morbidity and mortality associated with vein graft failure may be thereby reduced. Materials and methods Harvesting/arterial reconstruction. Fourteen mongrel dogs weighing 20 to 25 kg underwent bilateral femoral artery reconstruction with reversed cephalic veins. Animal care and surgicalprocedures complied with the "Principlesof Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication no. 80-23, revised 1978). Each animal received two femoral artery bypass grafts for which reversed cephalic vein segmentswere used. On one side, the vein graft was prepared with an "injury" techniqueand on the other sidewithour "optimal" technique. The injury technique consisted of incision directlyoverthe veinfollowed by ligationof tributaries, excision of the vein,and syringedistentionwith normal saline.The injury-prepared veins were then stored in saline at 37° C for 45 minutesbeforeimplantation.This techniqueuniformlydisrupts but does not denude endothelialcell integrity before grafting.' The optimal technique involved transcutaneous perivenous 671
The Journal of Thoracic and Cardiovascular Surgery
67 2 Quist. Haudenschild, LoGerfo
Fig. 1. Scanning electron microgr aph of a baseline vein harvested according to the "injury" protocol. There is significant damage to the intima characterized by cellular separation, cell lifting, and desquamation. Blebs of retracted endothelial cells extend into the lumen (original magn ification X1200). infiltration with papaverine solution followed by skin incision and repeat perivenous papaverine infiltration under direct vision. The tributaries were then ligated. The veins were distended in situ with warm (37 0 C) papaverine solution at a controlled pressure of 300 mm Hg and then stored in cold (4 0 C) papaverine solution for 45 minutes . The papaverine solution consisted of 60 mg papaverine, 2000 U heparin , and 12.5 gm albumin in 500 ml electrolyte solution (Plasma-Lyte A) . This leaves endotheli al cell structure 85% to 1000/0 intact with only occasional dropout. Further details on the injury and optimal techniques and their immediate effects on endothelial cell structure have been described previously.' Femoral arteries were reconstructed with standard vascular techniques, including heparin (100 U( kg), protamine sulfate, vessel loops, 6-0 Prolene sutures (Ethicon, Inc., Somerville, N.J .), and loupe magnification. On completion of the proximal anastomosis, an y residual vein was immersion-fixed in 4% formaldehyde and glutaraldehyde (4CF-I G) and submitted for light, scanning, and transmission electron microscopy. This documented baseline structure of the graft at implantation (Figs . I and 2). Animals were assigned by random allocation to be put to death at 3 and 24 hours, 7, 30, 60, 90, 150, and 432 da ys after implantation. When the animals were put to death, the arterial vessels proximal to and distal to the graft were cannulated without disturbing the graft or tissues overlying the graft. After
Fig. 2. Transmission electron micrograph of a canine cephalic vein prepared according to the optimal protocol. Endothelial cells are closely adherent with apposition of cell borders .Smooth muscle cells are morphologically intact a nd of the filament-rich cell phenotype without edema of the surrounding extracellular matrix (original magnification X4900) .
heparinization and under occlusion control, the grafts were cleared of blood by means of lactated Ringer's solution under pressure of gravity. This was immediately followed by flushing a nd distention of the graft in situ with 4CF-l Gat 300 mm Hg pressure. The skin and tissues overlying the graft were then incised. The distended fixed graft and adjacent vessels were removed in toto and further immersion-fixed under pressure for 24 hours before processing. This protocol for blood vessel fixation has previously been reported and found necessary in the prevention of fixation artifact as a result of vessel collapse and trauma resulting from dissection.F' All specimens for morphologic a nalysis were taken I ern from either anastomosis to ensure that the segments being analyzed were free of trauma resulting from operative instrumentation. Specimens for scanning electron microscopy were processed through slowly stirred ethanols, dried in a Polaron critical-point dryer (Bio-Rad, Microscience Division, Cambridge, Mass.), sputter-coa ted with gold palladium alloy, and viewedon a JEOL SEM 5 microscope (JEOL USA Inc., Peabody , Mass .) at 20
kY.
Specimens for scanning electron microscopy were postfixed with osmium, processed through graded ethanols, and impregnated with Epon 812 embedding medium . Ultrathin sections
Volume 103 Number 4 April 1992
Qualitative microscopy of implanted vein grafts
673
Fig. 3. Transmission electron micrographof an optimallyprepared veingraft after 7 days. Integrity of endothelial layerisevidenced by no area of denudationand closeapposition of cellmembranesbetweenadjacent endothelialcells (arrow). The smooth musclecell population is primarily filamentous (smc) in phenotype, with only a few metabolic organelles seen closeto the nucleus. An occasional organelle-rich cell was seen in the media (are). The extracellular matrix is composed of mature cross-linked collagen fibers (c).
werecut and stained with uranyl acetate and lead citrate and viewed on a Philips electron microscope (Philips Electronic Instruments Inc., Mahwah, N.J.).
Results Optimally prepared vein grafts. In the acute postoperative period (3 and 24 hours), vein grafts prepared according to the optimal protocol maintained endothelial cell confluence over the entire surface of the graft. There was no cellular, platelet, or proteinaceous material adherent to the intimal surface. Seven days after grafting, the anastomosis had healed. The smooth muscle cells remained primarily filament rich in phenotype without morphologic changes occurring in the extracellular matrix. Occasional organelle-rich phenotypes were seen but constituted less than 10% of the total smooth muscle cell population (Fig. 3). The perivascular capillaries were restored and in continuity with the surrounding tissues. Optimally prepared vein grafts excised at 30 days or
longer had a morphologic profile similar to the baseline segment, with confluent endothelium, filament-rich smooth muscle cells, and mature collagen in the extracellular matrix (Fig. 4). Injured vein grafts. Three hours after reconstruction, vein grafts prepared according to the injury protocol had moderate desquamation of endothelial cells. In areas not covered by remnant endothelium, platelet adhesion to denuded areas with white blood cell invasion of the graft wall could be seen. By 24 hours, a fine carpet of activated platelets and white blood cells could be seen adherent to the intimal surface of injured vein grafts (Fig. 5). The walls of the grafts at this time had a marked inflammatory response throughout their entirety. By 7 days, however, the lumina of injured grafts were relined by a morphologically intact endothelium without adherence of platelets or white cells. The media, however, had smooth muscle cells that were phenotypically mixed. Both transitional and organelle-rich smooth muscle cell phenotypes
674
Quist, Haudenschild, LoGerfo
Fig. 4. Transmission electron micrograph of an optimally prepared vein graft 428 daysafter reconstruction. The lumen is lined by endothelium. The media is composed of filament-rich smooth muscle cells with mature collagen in the extracellular matrix (original magnification X2800.) could be demonstrated (Fig. 6). Remodeling of the extracellular matrix from previously mature collagen fibers to amorphous material was also seen. By 30 days, the smooth muscle cells of initially injured vein grafts were 100% of the organelle-rich cell phenotype with a marked absence of the filament-rich cell phenotypes.The extracellular matrix was completelyamorphous in nature (Fig. 7). In vein grafts excised at intervals of 70 days or more, the lumina were fully lined by endothelial cells, media consisted of smooth muscle cells of the filamentrich cellphenotype,and the extracellular matrices showed mature collagen fibers (Fig. 8). Discussion Patency rates for reversed saphenous vein grafts have been reported between 500/0 and 86% at 5 years of followup according to life-table analysis.v" There is uniform agreement that, other than technical failures, acute thrombosis and delayed intimal hyperplasia are the two
The Journal of Thoracic and Cardiovascular Surgery
Fig, 5. Scanning electron micrograph ofan injured vein graft after 24 hours of arterial blood flow. A carpet of activated platelets and white blood cells can beseen over theentire graft surface. Overlying thegraftsurface are remnants ofendothelial cells (original magnification x2000). processes most closely linked with vein graft failure. Understanding the events that lead to these processes continues to be an area of active clinical and basicscience research. In 1981 we proposed that standard clinical protocols for vein graft harvesting resulted in endothelial cell injury.3.9 We characterized the effects of several graft preparation techniques that explored the cause-and-effect relationship of vasospasm, temperature, pressure, and fluid composition on vein graft integrity before implantation. We also developeda technique incorporating the use of transcutaneous perivenous papaverine infiltration to prevent venospasmand facilitate the preservationof vein graft endothelium and smooth musclecell integritybefore arterial reconstruction.The current study wasdesignedto incorporate the information learned in our previous studies to determine if there is a relationship betweenvein graft preparation technique and postimplantation vein graft abnormalities, that is, acute platelet adherence and intimal hyperplasia. Conversely, are these processes independent of preparation technique and the result of
Volume 103 Number 4 April 1992
Qualitative microscopy of implanted vein grafts
675
Fig. 6. Transmission electron micrograph ofan injured vein graftafter7 days. Endothelial integrity hasbeen reestablished. Thesmooth muscle cell population ismixed. In transitional SMC the predominant intracellular organelle present isrough endoplasmic reticulum engorged with amorphous material. Theextracellular matrix isless organized, with diminished mature collagen and amorphous material interspersed between individual smooth muscle cells. TC and ORC, transitional andorganelle-rich smooth muscle cell phenotypes; EC, endothelial cells (original magnification X2800). obligatory events associated with transplantation to the arterial tree, a process some have referred to as "arterialization". Acute thrombosis in the vascular tree has long been associated with antecedent endothelial cell "injury." Vasospasm, high pressure, and saline storage solutions have each been independently associated with morphologic injury to endothelialcellsin veingrafts.': 4.10. II The thrombogenicity of such a conduit with injured endothelial cells will be determined by the degree of cellular damage, the amount of exposed subendothelialcollagen, and the velocity of blood through the graft. 12 The magnitude of injury will result in localized platelet and thrombus deposition or even complete graft thrombosis. The current study has established two important phenomenaof optimally prepared veingrafts. First veingraft endothelium, if initially intact after storage, remains morphologically intact despite being interposed in the arterial tree with its associated higher shear forces. Second, not only is morphologic integrity maintained, but function also appears to be preserved. Platelets do not adhere to the intact intima, and there were no areas of localized thrombus deposition. This is in marked contrast to pair-matchedcontralateral injured veingrafts in which platelet adhesion and activation occur early in the postoperative period. Although these events have been well documented for injured arterial segments, only recently haveour work and that of others confirmedthis relation-
ship of initial veingraft integrity with events after in vivo arterial reconstruction. It is therefore possible to conclude that preservation of vein graft integrity before implantation is beneficial and should decrease the prevalence of acute vein graft failures. The proliferationof smooth musclecellsin culture and in traumatized arterial segments has similarly been thoroughly studied.I3, 14 Indeed, smooth muscle cell proliferation is believed to be a pivotal event in arterial wall lesions.P Growth factors for smooth muscle cells have been found in platelets," endothelium.l? and smooth musclecellsthemselves. 18 Whether theseendogenousand exogenous growth factors playa role in vein graft intimal hyperplasia is not clear. Much controversysurrounds the mechanism of intimal hyperplasia within vein grafts and whether it is an inevitablehealing and adaptive response to arterial pressure and flow (unavoidable) or a physiologic response to acute and chronic injury (avoidable). Szilagyi,Elliot,and Hageman, 19in reviewing clinicalvein graft failures,designated intimal hyperplasia as the most common lesion predisposing to vein graft failure in the delayed postoperative period. That intimal hyperplasia occurs in veingrafts has been repeatedlydemonstrated in both the clinical and research literature. This study demonstrates a direct relationship between the morphologic integrity of the vein graft before grafting and delayed morphologic events.The smooth muscle cells of the injured vein grafts underwent a distinct phe-
676
Quist. Haudenschild, LoGerfo
:;r;~,@~·.':;;}i:t.~~iit~
Fig. 7. Transmission electron micrograph of an injury-prepared vein graft at 30 days. Endothelium lines the lumen and formsclosely apposed cell-to-cell junctions. The media is composedonly of organelle-rich smooth musclecells. Rough endoplasmic reticulum is the predominant intracellular organelle seen. The extracellular matrix iscomposed of amorphousstaining material (original magnification X4560).
notypic change from contractile (filament-rich cells) to metabolic phenotypes (organelle-rich cells). This phenotypic change has been associated with increased synthesis of extracellular matrix proteins and is thought by some to be necessary for smooth muscle cell proliferation to occur.i" In a related study, we have performed morphometric analysis comparing optimally prepared versus injury-prepared vein grafts. That study showed injuryprepared grafts have thicker walls for all time periods evaluated when compared with pair-matched optimally prepared vein grafts.I' Whether this is a function of increased cell number or increased matrix is currently under investigation. Thus, in this defined and controlled model, the natural history of initial vein graft injury is early (24 hours to 7 days) healing of the intima from remnant endothelium with delayed smooth muscle cell modulation and remodeling of the extracellular matrix. Late morphologic changes occurring in vein grafts ini-
The Journal of Thoracic and Cardiovascular Surgery
tially injured include smooth muscle cells that are primarily filamentous in phenotype and extracellular matrices of mature collagen fibers. Thus it appears that a dynamic phase of healing and remodeling of the vein graft wall occurs after initial injury. In this model of limited vein graft injury, morphologic changes within the smooth muscle cell population mimic the lesions of occlusive intimal hyperplasia. However in this study, occlusive lesions did not occur; rather, resumption of filamentous phenotype resulted. What stimuli are responsible for this initial phenotypic change is still poorly understood, although various growth factors, mitogens, and hemodynamic forces have been implicated. What causes a hyperplastic lesion to progress, or, in this case, revert to its filamentous phenotype remains an area for active research. This model of limited graft injury provides a reproducible window at 30 days within which to study the dynamics of smooth muscle cells in a transitional state. Optimally prepared vein grafts maintained smooth muscle cell structure in the acute, delayed, and late postoperative settings. Synthetic phenotypes were rare to absent. Extracellular matrices in optimally prepared vein grafts did not undergo gross remodeling as seen in injuryprepared vein grafts. Several possible mechanisms responsible for control of smooth muscle cell growth have been investigated previously by other authors. Studies of extracellular matrix in other biologic systems and specifically in developmental biology have shown profound effects of the extracellular matrix on cell structure and function.F Recent studies with smooth muscle cell growth on varied substrates of collagen, fibronectin, and other matrix proteins and proteoglycans have shown this dependent relationship. Thus arterial wall biology and, in this case, vein graft wall biology must now be understood in terms of blood rheology, endothelial cell, smooth muscle cell, and extracellular matrix interactions. It is therefore concluded from this study that optimal preservation of vein grafts before arterial reconstruction has a profound and beneficial effect on vein graft wall structure after grafting. This favorable impact on vein grafts should result in a decrease in prevalence of both acute thromboses and delayed intimal hyperplasia, and it should ultimately result in improved patency rates in vascular procedures for which autologous vein grafts are used. REFERENCES I. Fogle MA, Whittemore AD, Couch NP, Mannick JA. A comparison of in situ and reversed saphenousvein grafts for infrainguinal reconstruction. J Vase Surg 1987;5:46-52 . 2. Grondon CM, Campeau L, Lesperance J, Enjalbest M, Bourassa M. Comparisonof late changesin internal mammary artery and saphenousvein grafts in two consecutive
Volume 103 Number 4
April 1992
Qualitative microscopy of implanted vein grafts
677
Fig. 8. Transmission electron micrograph of a 428-day vein graft initially prepared according to the injury protocol. Endothelial cells line the lumen. The media is composed of filament-rich smooth muscle cells exclusively. The extracellular matrix is composed of cross-linked collagen (original magnification X2800).
3.
4.
5.
6.
7.
8.
9.
10.
11.
seriesof patients 10 years after operation. Circulation 1984; 70:208-12. LoGerfo FW, Quist WC, Crawshaw HM, Haudenschild Ce. An improved technique for preservation of endothelial morphology in vein grafts. Surgery 1981;90:1015-24. LoGerfo FW, Quist WC, Cantelmo NL, Haudenschild Ce. Integrity of vein grafts as a function of initial intimal and medial preservation. Circulation 1983;68:117-24. Haudenschild CC, Baumgartner HR, Studer A. Significance of fixation procedure for preservation of arteries. Experientia 1972;28:828-31. Cheseboro JH, Fuster V, Elueback C, et al. Effect of dipyridamole and aspirin on late vein graft patency after coronary bypass operations. N Engl J Med 1984;310:209-14. Taylor LM, Phinney ES, Porter JM. Present status of reversed vein grafts for lower extremity revascularization. J Vase Surg 1986;3:288-97. Taylor LM, Edwards JM, Phinney ES, Porter JM. Reversed vein bypass to infra popliteal arteries: modern results are superior to or equivalent to in-situ bypass for patency and for vein utilization. Ann Surg 1987;205: 90-7. Haudenschild CC, Quist WC, Gould KE, LoGerfo FW. Protection of endothelium in vessel segments excised for grafting. Circulation 1981;64:101-7. Stanley JC, Sottiurai V, Fry R, Fry W. Comparative evaluation of vein graft preparation media: electron and light microscopic studies. J Surg Res 1975;13:235-46. Bonchek Ll. Prevention of endothelial damage during preparation of saphenous vein for bypass grafting. J THORAC CARDIOVASC SURG 1980;79:911-5.
12. Baumgartner H, Turitto V, Weiss H. Effect of shear rate on platelet interaction with subendothelium in citrated and natural blood. II. Relationship among platelet adhesion, thrombus dimensions and fibrin formation. J Lab Clin Med 1980;95:208-21. 13. Chamley-Campbell JH, Campbell GR, Ross R. Review. Smooth muscle in culture. Physiol Rev 1979;5:1-61. 14. Clowes A, Reidy M, Clowes M. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 1983;49:327-33. 15. Ross R, Glomset JA. Atherosclerosis and the arterial smooth muscle cell. Science 1973;180:1332-9. 16. Heldin CH, Wasteson A, Westermark B. Review platelet derived growth factor. Mol Cell Endocrino1 1985;39:16987. 17. Collins T, Puber J, Gibrone M, et a1. Cultured human endothelial cells express platelet derived growth factor A chain. Am J Pathol 1987;126:7-12. 18. Nilsson J, Sjolund M, Palimberg L, Thybert J, Heldin L. Arterial smooth muscle cells in primary culture produce a platelet derived growth factor-like protein. Proc N atl Acad Sci USA 1985;82:4418-22. 19. Szilagyi D, Elliot J, Hageman J. Biologicfate of autogenous vein implants as arterial substitutes. Ann Surg 1973; 178:232-44. 20. Chamley-Campbell J, Campbell F, Ross R. Phenotype dependent response of cultured aortic smooth muscle to serum mitogens. J Cell Bioi 1981;89:379-83. 21. Quist We. Autologous vein as an arterial conduit: experimental and clinical studies [Dissertation] Boston, Mass.: Boston University, 1988.
William C. Quist, MD, PhD, Christian C. Haudenschild, MD, and Frank W. LoGerfo, MD, Boston, Mass.
h e saphenous vein continues to be the optimal graft for infrapopliteal and complete cardiac revascularizations.'- 2 Despite this, a significant number of vein graft failures occur in the acute and delayed postoperative period. Endothelial cell injury and smooth muscle cell proliferation appear to be important factors leading to these vein graft failures. Early studies from this laboratory implicated harvesting techniques as a mechanism for both endothelial and smooth muscle cell damage before arterial reconstruction. Postulating a relationship between graft integrity at the time of implantation and subsequent graft function, we developed a technique that led to optimal preservation of both endothelial and smooth muscle cells before arterial reconstruction.' Graft integrity was best preserved with transcutaneous perivenous papaverine, controlled temperature and pressure during distention and storage, and maintenance of distention during the storage and reconstructive pericds.l 4 From the Divisionof Vascular Surgery, Harvard Medical School, New England Deaconess Hospital, and the Mallory Institute of Pathology, Boston University School of Medicine, Boston, Mass. Received for publication Aug. 22, 1990. Accepted for publication Feb. 7, 1991. Address for reprints: Frank W. LoGerfo, MD, New England Deaconess Hospital, 110 Francis St., Boston, MA 02215.
12/1/29023
The purpose of this study was to use qualitative microscopy to compare serially the effects of vein graft preparation techniques on the subsequent morphologic events after arterial reconstruction. If vein graft preparation technique has an impact on morphologic events after arterial reconstruction, a significant number of vein graft failures may be prevented and the morbidity and mortality associated with vein graft failure may be thereby reduced. Materials and methods Harvesting/arterial reconstruction. Fourteen mongrel dogs weighing 20 to 25 kg underwent bilateral femoral artery reconstruction with reversed cephalic veins. Animal care and surgicalprocedures complied with the "Principlesof Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication no. 80-23, revised 1978). Each animal received two femoral artery bypass grafts for which reversed cephalic vein segmentswere used. On one side, the vein graft was prepared with an "injury" techniqueand on the other sidewithour "optimal" technique. The injury technique consisted of incision directlyoverthe veinfollowed by ligationof tributaries, excision of the vein,and syringedistentionwith normal saline.The injury-prepared veins were then stored in saline at 37° C for 45 minutesbeforeimplantation.This techniqueuniformlydisrupts but does not denude endothelialcell integrity before grafting.' The optimal technique involved transcutaneous perivenous 671
The Journal of Thoracic and Cardiovascular Surgery
67 2 Quist. Haudenschild, LoGerfo
Fig. 1. Scanning electron microgr aph of a baseline vein harvested according to the "injury" protocol. There is significant damage to the intima characterized by cellular separation, cell lifting, and desquamation. Blebs of retracted endothelial cells extend into the lumen (original magn ification X1200). infiltration with papaverine solution followed by skin incision and repeat perivenous papaverine infiltration under direct vision. The tributaries were then ligated. The veins were distended in situ with warm (37 0 C) papaverine solution at a controlled pressure of 300 mm Hg and then stored in cold (4 0 C) papaverine solution for 45 minutes . The papaverine solution consisted of 60 mg papaverine, 2000 U heparin , and 12.5 gm albumin in 500 ml electrolyte solution (Plasma-Lyte A) . This leaves endotheli al cell structure 85% to 1000/0 intact with only occasional dropout. Further details on the injury and optimal techniques and their immediate effects on endothelial cell structure have been described previously.' Femoral arteries were reconstructed with standard vascular techniques, including heparin (100 U( kg), protamine sulfate, vessel loops, 6-0 Prolene sutures (Ethicon, Inc., Somerville, N.J .), and loupe magnification. On completion of the proximal anastomosis, an y residual vein was immersion-fixed in 4% formaldehyde and glutaraldehyde (4CF-I G) and submitted for light, scanning, and transmission electron microscopy. This documented baseline structure of the graft at implantation (Figs . I and 2). Animals were assigned by random allocation to be put to death at 3 and 24 hours, 7, 30, 60, 90, 150, and 432 da ys after implantation. When the animals were put to death, the arterial vessels proximal to and distal to the graft were cannulated without disturbing the graft or tissues overlying the graft. After
Fig. 2. Transmission electron micrograph of a canine cephalic vein prepared according to the optimal protocol. Endothelial cells are closely adherent with apposition of cell borders .Smooth muscle cells are morphologically intact a nd of the filament-rich cell phenotype without edema of the surrounding extracellular matrix (original magnification X4900) .
heparinization and under occlusion control, the grafts were cleared of blood by means of lactated Ringer's solution under pressure of gravity. This was immediately followed by flushing a nd distention of the graft in situ with 4CF-l Gat 300 mm Hg pressure. The skin and tissues overlying the graft were then incised. The distended fixed graft and adjacent vessels were removed in toto and further immersion-fixed under pressure for 24 hours before processing. This protocol for blood vessel fixation has previously been reported and found necessary in the prevention of fixation artifact as a result of vessel collapse and trauma resulting from dissection.F' All specimens for morphologic a nalysis were taken I ern from either anastomosis to ensure that the segments being analyzed were free of trauma resulting from operative instrumentation. Specimens for scanning electron microscopy were processed through slowly stirred ethanols, dried in a Polaron critical-point dryer (Bio-Rad, Microscience Division, Cambridge, Mass.), sputter-coa ted with gold palladium alloy, and viewedon a JEOL SEM 5 microscope (JEOL USA Inc., Peabody , Mass .) at 20
kY.
Specimens for scanning electron microscopy were postfixed with osmium, processed through graded ethanols, and impregnated with Epon 812 embedding medium . Ultrathin sections
Volume 103 Number 4 April 1992
Qualitative microscopy of implanted vein grafts
673
Fig. 3. Transmission electron micrographof an optimallyprepared veingraft after 7 days. Integrity of endothelial layerisevidenced by no area of denudationand closeapposition of cellmembranesbetweenadjacent endothelialcells (arrow). The smooth musclecell population is primarily filamentous (smc) in phenotype, with only a few metabolic organelles seen closeto the nucleus. An occasional organelle-rich cell was seen in the media (are). The extracellular matrix is composed of mature cross-linked collagen fibers (c).
werecut and stained with uranyl acetate and lead citrate and viewed on a Philips electron microscope (Philips Electronic Instruments Inc., Mahwah, N.J.).
Results Optimally prepared vein grafts. In the acute postoperative period (3 and 24 hours), vein grafts prepared according to the optimal protocol maintained endothelial cell confluence over the entire surface of the graft. There was no cellular, platelet, or proteinaceous material adherent to the intimal surface. Seven days after grafting, the anastomosis had healed. The smooth muscle cells remained primarily filament rich in phenotype without morphologic changes occurring in the extracellular matrix. Occasional organelle-rich phenotypes were seen but constituted less than 10% of the total smooth muscle cell population (Fig. 3). The perivascular capillaries were restored and in continuity with the surrounding tissues. Optimally prepared vein grafts excised at 30 days or
longer had a morphologic profile similar to the baseline segment, with confluent endothelium, filament-rich smooth muscle cells, and mature collagen in the extracellular matrix (Fig. 4). Injured vein grafts. Three hours after reconstruction, vein grafts prepared according to the injury protocol had moderate desquamation of endothelial cells. In areas not covered by remnant endothelium, platelet adhesion to denuded areas with white blood cell invasion of the graft wall could be seen. By 24 hours, a fine carpet of activated platelets and white blood cells could be seen adherent to the intimal surface of injured vein grafts (Fig. 5). The walls of the grafts at this time had a marked inflammatory response throughout their entirety. By 7 days, however, the lumina of injured grafts were relined by a morphologically intact endothelium without adherence of platelets or white cells. The media, however, had smooth muscle cells that were phenotypically mixed. Both transitional and organelle-rich smooth muscle cell phenotypes
674
Quist, Haudenschild, LoGerfo
Fig. 4. Transmission electron micrograph of an optimally prepared vein graft 428 daysafter reconstruction. The lumen is lined by endothelium. The media is composed of filament-rich smooth muscle cells with mature collagen in the extracellular matrix (original magnification X2800.) could be demonstrated (Fig. 6). Remodeling of the extracellular matrix from previously mature collagen fibers to amorphous material was also seen. By 30 days, the smooth muscle cells of initially injured vein grafts were 100% of the organelle-rich cell phenotype with a marked absence of the filament-rich cell phenotypes.The extracellular matrix was completelyamorphous in nature (Fig. 7). In vein grafts excised at intervals of 70 days or more, the lumina were fully lined by endothelial cells, media consisted of smooth muscle cells of the filamentrich cellphenotype,and the extracellular matrices showed mature collagen fibers (Fig. 8). Discussion Patency rates for reversed saphenous vein grafts have been reported between 500/0 and 86% at 5 years of followup according to life-table analysis.v" There is uniform agreement that, other than technical failures, acute thrombosis and delayed intimal hyperplasia are the two
The Journal of Thoracic and Cardiovascular Surgery
Fig, 5. Scanning electron micrograph ofan injured vein graft after 24 hours of arterial blood flow. A carpet of activated platelets and white blood cells can beseen over theentire graft surface. Overlying thegraftsurface are remnants ofendothelial cells (original magnification x2000). processes most closely linked with vein graft failure. Understanding the events that lead to these processes continues to be an area of active clinical and basicscience research. In 1981 we proposed that standard clinical protocols for vein graft harvesting resulted in endothelial cell injury.3.9 We characterized the effects of several graft preparation techniques that explored the cause-and-effect relationship of vasospasm, temperature, pressure, and fluid composition on vein graft integrity before implantation. We also developeda technique incorporating the use of transcutaneous perivenous papaverine infiltration to prevent venospasmand facilitate the preservationof vein graft endothelium and smooth musclecell integritybefore arterial reconstruction.The current study wasdesignedto incorporate the information learned in our previous studies to determine if there is a relationship betweenvein graft preparation technique and postimplantation vein graft abnormalities, that is, acute platelet adherence and intimal hyperplasia. Conversely, are these processes independent of preparation technique and the result of
Volume 103 Number 4 April 1992
Qualitative microscopy of implanted vein grafts
675
Fig. 6. Transmission electron micrograph ofan injured vein graftafter7 days. Endothelial integrity hasbeen reestablished. Thesmooth muscle cell population ismixed. In transitional SMC the predominant intracellular organelle present isrough endoplasmic reticulum engorged with amorphous material. Theextracellular matrix isless organized, with diminished mature collagen and amorphous material interspersed between individual smooth muscle cells. TC and ORC, transitional andorganelle-rich smooth muscle cell phenotypes; EC, endothelial cells (original magnification X2800). obligatory events associated with transplantation to the arterial tree, a process some have referred to as "arterialization". Acute thrombosis in the vascular tree has long been associated with antecedent endothelial cell "injury." Vasospasm, high pressure, and saline storage solutions have each been independently associated with morphologic injury to endothelialcellsin veingrafts.': 4.10. II The thrombogenicity of such a conduit with injured endothelial cells will be determined by the degree of cellular damage, the amount of exposed subendothelialcollagen, and the velocity of blood through the graft. 12 The magnitude of injury will result in localized platelet and thrombus deposition or even complete graft thrombosis. The current study has established two important phenomenaof optimally prepared veingrafts. First veingraft endothelium, if initially intact after storage, remains morphologically intact despite being interposed in the arterial tree with its associated higher shear forces. Second, not only is morphologic integrity maintained, but function also appears to be preserved. Platelets do not adhere to the intact intima, and there were no areas of localized thrombus deposition. This is in marked contrast to pair-matchedcontralateral injured veingrafts in which platelet adhesion and activation occur early in the postoperative period. Although these events have been well documented for injured arterial segments, only recently haveour work and that of others confirmedthis relation-
ship of initial veingraft integrity with events after in vivo arterial reconstruction. It is therefore possible to conclude that preservation of vein graft integrity before implantation is beneficial and should decrease the prevalence of acute vein graft failures. The proliferationof smooth musclecellsin culture and in traumatized arterial segments has similarly been thoroughly studied.I3, 14 Indeed, smooth muscle cell proliferation is believed to be a pivotal event in arterial wall lesions.P Growth factors for smooth muscle cells have been found in platelets," endothelium.l? and smooth musclecellsthemselves. 18 Whether theseendogenousand exogenous growth factors playa role in vein graft intimal hyperplasia is not clear. Much controversysurrounds the mechanism of intimal hyperplasia within vein grafts and whether it is an inevitablehealing and adaptive response to arterial pressure and flow (unavoidable) or a physiologic response to acute and chronic injury (avoidable). Szilagyi,Elliot,and Hageman, 19in reviewing clinicalvein graft failures,designated intimal hyperplasia as the most common lesion predisposing to vein graft failure in the delayed postoperative period. That intimal hyperplasia occurs in veingrafts has been repeatedlydemonstrated in both the clinical and research literature. This study demonstrates a direct relationship between the morphologic integrity of the vein graft before grafting and delayed morphologic events.The smooth muscle cells of the injured vein grafts underwent a distinct phe-
676
Quist. Haudenschild, LoGerfo
:;r;~,@~·.':;;}i:t.~~iit~
Fig. 7. Transmission electron micrograph of an injury-prepared vein graft at 30 days. Endothelium lines the lumen and formsclosely apposed cell-to-cell junctions. The media is composedonly of organelle-rich smooth musclecells. Rough endoplasmic reticulum is the predominant intracellular organelle seen. The extracellular matrix iscomposed of amorphousstaining material (original magnification X4560).
notypic change from contractile (filament-rich cells) to metabolic phenotypes (organelle-rich cells). This phenotypic change has been associated with increased synthesis of extracellular matrix proteins and is thought by some to be necessary for smooth muscle cell proliferation to occur.i" In a related study, we have performed morphometric analysis comparing optimally prepared versus injury-prepared vein grafts. That study showed injuryprepared grafts have thicker walls for all time periods evaluated when compared with pair-matched optimally prepared vein grafts.I' Whether this is a function of increased cell number or increased matrix is currently under investigation. Thus, in this defined and controlled model, the natural history of initial vein graft injury is early (24 hours to 7 days) healing of the intima from remnant endothelium with delayed smooth muscle cell modulation and remodeling of the extracellular matrix. Late morphologic changes occurring in vein grafts ini-
The Journal of Thoracic and Cardiovascular Surgery
tially injured include smooth muscle cells that are primarily filamentous in phenotype and extracellular matrices of mature collagen fibers. Thus it appears that a dynamic phase of healing and remodeling of the vein graft wall occurs after initial injury. In this model of limited vein graft injury, morphologic changes within the smooth muscle cell population mimic the lesions of occlusive intimal hyperplasia. However in this study, occlusive lesions did not occur; rather, resumption of filamentous phenotype resulted. What stimuli are responsible for this initial phenotypic change is still poorly understood, although various growth factors, mitogens, and hemodynamic forces have been implicated. What causes a hyperplastic lesion to progress, or, in this case, revert to its filamentous phenotype remains an area for active research. This model of limited graft injury provides a reproducible window at 30 days within which to study the dynamics of smooth muscle cells in a transitional state. Optimally prepared vein grafts maintained smooth muscle cell structure in the acute, delayed, and late postoperative settings. Synthetic phenotypes were rare to absent. Extracellular matrices in optimally prepared vein grafts did not undergo gross remodeling as seen in injuryprepared vein grafts. Several possible mechanisms responsible for control of smooth muscle cell growth have been investigated previously by other authors. Studies of extracellular matrix in other biologic systems and specifically in developmental biology have shown profound effects of the extracellular matrix on cell structure and function.F Recent studies with smooth muscle cell growth on varied substrates of collagen, fibronectin, and other matrix proteins and proteoglycans have shown this dependent relationship. Thus arterial wall biology and, in this case, vein graft wall biology must now be understood in terms of blood rheology, endothelial cell, smooth muscle cell, and extracellular matrix interactions. It is therefore concluded from this study that optimal preservation of vein grafts before arterial reconstruction has a profound and beneficial effect on vein graft wall structure after grafting. This favorable impact on vein grafts should result in a decrease in prevalence of both acute thromboses and delayed intimal hyperplasia, and it should ultimately result in improved patency rates in vascular procedures for which autologous vein grafts are used. REFERENCES I. Fogle MA, Whittemore AD, Couch NP, Mannick JA. A comparison of in situ and reversed saphenousvein grafts for infrainguinal reconstruction. J Vase Surg 1987;5:46-52 . 2. Grondon CM, Campeau L, Lesperance J, Enjalbest M, Bourassa M. Comparisonof late changesin internal mammary artery and saphenousvein grafts in two consecutive
Volume 103 Number 4
April 1992
Qualitative microscopy of implanted vein grafts
677
Fig. 8. Transmission electron micrograph of a 428-day vein graft initially prepared according to the injury protocol. Endothelial cells line the lumen. The media is composed of filament-rich smooth muscle cells exclusively. The extracellular matrix is composed of cross-linked collagen (original magnification X2800).
3.
4.
5.
6.
7.
8.
9.
10.
11.
seriesof patients 10 years after operation. Circulation 1984; 70:208-12. LoGerfo FW, Quist WC, Crawshaw HM, Haudenschild Ce. An improved technique for preservation of endothelial morphology in vein grafts. Surgery 1981;90:1015-24. LoGerfo FW, Quist WC, Cantelmo NL, Haudenschild Ce. Integrity of vein grafts as a function of initial intimal and medial preservation. Circulation 1983;68:117-24. Haudenschild CC, Baumgartner HR, Studer A. Significance of fixation procedure for preservation of arteries. Experientia 1972;28:828-31. Cheseboro JH, Fuster V, Elueback C, et al. Effect of dipyridamole and aspirin on late vein graft patency after coronary bypass operations. N Engl J Med 1984;310:209-14. Taylor LM, Phinney ES, Porter JM. Present status of reversed vein grafts for lower extremity revascularization. J Vase Surg 1986;3:288-97. Taylor LM, Edwards JM, Phinney ES, Porter JM. Reversed vein bypass to infra popliteal arteries: modern results are superior to or equivalent to in-situ bypass for patency and for vein utilization. Ann Surg 1987;205: 90-7. Haudenschild CC, Quist WC, Gould KE, LoGerfo FW. Protection of endothelium in vessel segments excised for grafting. Circulation 1981;64:101-7. Stanley JC, Sottiurai V, Fry R, Fry W. Comparative evaluation of vein graft preparation media: electron and light microscopic studies. J Surg Res 1975;13:235-46. Bonchek Ll. Prevention of endothelial damage during preparation of saphenous vein for bypass grafting. J THORAC CARDIOVASC SURG 1980;79:911-5.
12. Baumgartner H, Turitto V, Weiss H. Effect of shear rate on platelet interaction with subendothelium in citrated and natural blood. II. Relationship among platelet adhesion, thrombus dimensions and fibrin formation. J Lab Clin Med 1980;95:208-21. 13. Chamley-Campbell JH, Campbell GR, Ross R. Review. Smooth muscle in culture. Physiol Rev 1979;5:1-61. 14. Clowes A, Reidy M, Clowes M. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 1983;49:327-33. 15. Ross R, Glomset JA. Atherosclerosis and the arterial smooth muscle cell. Science 1973;180:1332-9. 16. Heldin CH, Wasteson A, Westermark B. Review platelet derived growth factor. Mol Cell Endocrino1 1985;39:16987. 17. Collins T, Puber J, Gibrone M, et a1. Cultured human endothelial cells express platelet derived growth factor A chain. Am J Pathol 1987;126:7-12. 18. Nilsson J, Sjolund M, Palimberg L, Thybert J, Heldin L. Arterial smooth muscle cells in primary culture produce a platelet derived growth factor-like protein. Proc N atl Acad Sci USA 1985;82:4418-22. 19. Szilagyi D, Elliot J, Hageman J. Biologicfate of autogenous vein implants as arterial substitutes. Ann Surg 1973; 178:232-44. 20. Chamley-Campbell J, Campbell F, Ross R. Phenotype dependent response of cultured aortic smooth muscle to serum mitogens. J Cell Bioi 1981;89:379-83. 21. Quist We. Autologous vein as an arterial conduit: experimental and clinical studies [Dissertation] Boston, Mass.: Boston University, 1988.