- Email: [email protected]
Intimal changes in arteriovenous bypass grafts
Intimal changes in arteriovenous bypass grafts Effects of varying the angle of implantation at the proximal anastomosis and of producing stenosis in the distal runoff artery Autologous arteriovenous bypass grafts were constructed between external iliac arteries in 20 mongrel dogs to determine the development of intimal fibromuscular hyperplasia. Animals received grafts at 55, 90 or 120 degree angles relative to the proximal anastomosis. Five animals with 90 degree angle grafts were also subjected to 80 per cent stenosis in the runoff artery. All grafts showed the development of a fihromuscular tunica intima. Qualitative or quantitative intimal differences were not observed among grafts placed at 90 degree angles, 90 degree angles with runoff artery stenosis, or 120 degree angles. Crafts placed at 55 degree angles relative to the proximal anastomosis developed a consistently thicker fihromuscular layer in the tunica intima when compared to the other groups. These changes which were segmental in nature, occurred in the proximal and distal grafts and over the distal anastomosis.
M. G. Bond, Ph.D.,* J. R. Hostetler, Ph.D.,* P. E. Karayannacos, M.D.,** J. C. Geer, M.D.,*** and J. S. Vasko, M.D., ** Columbus, Ohio
Zxutologus vein bypass surgery for the treatment of obliterative arterial disease involving coronary and peripheral arteries is being performed with increasing frequency. Late failure of venous grafts has been reported in aorta-coronary bypasses1-10 and arterial bypasses of the lower extremities.11-14 These vein grafts often show the presence of a subendothelial fibromuscular layer which may effectively reduce lumen diameter. Various etiological factors suggested to explain this phenomenon have included angulation of the proximal anastomosis,15'16 turbulence in the area
of intact valves,2 trauma to the vein graft during the surgical procedure,15 and the presence of obstructive lesions in the arterial outflow tract.3 The purpose of this investigation is to determine what changes occur in the thickness and distribution of the subendothelial fibromuscular layer that result from (1) different angles of implantation at the proximal anastomosis, (2) stenosis in the distal runoff artery, and (3) the presence of valves within the reversed vein graft.
From the Departments of Anatomy, Pathology, and Surgery, The Ohio State University, Columbus, Ohio 43210. Supported in part by National Heart and Lung Institute Grant HL11897, Harker's Fund Grant, and Central Ohio Heart Chapter Grant 3554-A1.
Twenty male and female mongrel dogs, weighing between 20 and 25 kilograms, received autologous femoral vein bypass grafts between the external iliac arteries. Animals were equally divided into four groups based upon the angulation of the proximal anastomosis and the presence of 80 per cent stenosis in the runoff artery (Fig. 1). Group I animals received grafts at a 90 degree angle relative to the proximal artery, Group II animals received 90 degree angle grafts besides being subjected to an 80 per cent stenosis in the runoff artery, Group III animals received grafts at an acute angle (55 degrees ± 10 degrees), and Group IV animals re-
Received for publication Dec. 24, 1975. Accepted for publication Jan. 29, 1976. Address for reprints: M. G. Bond, Ph.D., Department of Comparative Medicine, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, N. C. 27103. *Department of Anatomy. "Division of Thoracic Surgery, Department of Surgery. ***Department of Pathology.
Methods
907
The Journal of Thoracic and Cardiovascular Surgery
9 0 8 Bond et al.
Aorta Right external iliac artery
Stenosis AV Graft Group IT, 90°angle+ 8 0 % stenosis
Group I , 90° angle
Group M, 55° angle Group W,
Fig. 1. Diagram of experimental model. AV, Arteriovenous.
ceived grafts at an obtuse angle (120 degrees ± 10). Thirty minutes before the operation the animals were premedicated with acepromazine (0.55 mg. per kilogram) and atropine (0.05 mg. per kilogram). Anesthesia was initiated with sodium thiopental (20 mg. per kilogram). An endotracheal tube was placed and connected to a volume-controlled respirator. Anesthesia was maintained with a mixture of nitrous oxide, halothane, and oxygen. A low midline laparotomy was performed under sterile conditions and the retroperitoneal space was entered distal to the aortic bifurcation. The external iliac arteries were mobilized and umbilical tapes placed around them. A skin incision was made over the right inguinocrural region and the femoral vein identified and mobilized over a length of 10 cm. The vein was dissected and handled with great care to avoid unnecessary injury to the vessel wall. Tributaries were ligated with 6-0 polypropylene monofilament suture material. The section of femoral vein was removed and flushed with heparinized saline, with special care taken to avoid overdistention. The vein segment was placed in saline at room temperature
until the femoral arteries were prepared for anastomosis. A segment of right external iliac artery was isolated between two atraumatic vascular clamps and a longitudinal arteriotomy was performed on the medial aspect of the artery. An end-to-side anastomosis was accomplished between the iliac artery and the distal end of the vein graft by use of a continuous suture with 6-0 polypropylene monofilament. At the completion of the anastomosis, the graft was cross-clamped near the proximal end and flow was restored through the right external iliac artery. The graft was passed through a surgical opening in the mesorectum and brought into apposition with the left external iliac artery, where a similar vascular anastomosis was performed. All clamps were removed and the left external iliac artery was twice ligated with 2-0 silk proximal to the anastomotic site. Blood flow was now from the right external iliac artery, designated the proximal artery, through the graft to the left external iliac artery, now designated the distal artery. The mesorectum was reconstructed with interrupted
Volume 71 Number 6 June, 1976
Intimal proliferation in femoral bypass grafts
3-0 catgut stitches. The peritoneum was closed with continuous 2-0 chromic catgut and the linea alba apposed with interrupted 3-0 Dacron. Skin was closed with Vetafil, a special veterinary suture material. Elapsed time between removal of the vein and completion of the graft anastomosis was approximately 30 minutes. Distances of the anastomotic sites from the aortic bifurcation and shapes of the proximal and distal ends of the vein graft varied depending on the angle between the proximal artery and the graft. The animals received antibiotic therapy with penicillin and streptomycin for one week, and sutures were removed on the tenth postoperative day. Ten days postoperatively, 5 animals with 90 degree angle anastomoses were reoperated upon. Under the same anesthetic and sterile surgical conditions, a low left paramedian incision was made. The left distal external iliac artery was identified and mobilized. A 60 per cent constriction in diameter of the artery, corresponding to an 80 per cent reduction in crosssectional area of the vessel, was achieved by a circumferential Mersilene tape 5 mm. wide. The abdomen was then closed in the described manner. All animals received postoperative treatment similar to that after the initial operation. Six months from the time of the original operation animals from all groups were again operated upon, under sodium nembutal anesthesia. Through a midline laparotomy the abdominal aorta, external iliac arteries, and the vein graft were dissected. The abdominal aorta was cross-clamped and a large cannula inserted into the vessel distal to the clamp. The posterior extremities and pelvis were perfused in situ under a pressure of 100 mm. Hg with 4° C. normal saline followed by a 3 per cent glutaraldehyde-formaldehyde fixing solution. Specimens were washed in 0.1 M P0 4 buffer solution (pH 7.4) at room temperature for 24 to 48 hours. The diameter of the graft was measured at operation with a caliper and the true diameter calculated after subtracting the thickness of the vein wall. After fixation, the graft was removed and opened longitudinally. Internal circumference was measured with a micrometer. The diameter was derived from the formula: circumference =
2TTR.
A 25 per cent shrinkage from the fixation process was considered. Representative sections for light and electron microscopic evaluation were taken from the proximal anastomosis (A), proximal graft (B), midgraft (C),
909
Table I. One-way analysis of variance for intimal thickening Region of graft A B C D E
Source of variation Between Between Between Between Between
Degrees of freedom
F
Level of significance
3 3 3 3 3
2.206 15.977 2.595 11.428 10.076
NS* p < O.Olt NS* p < 0.01$ p < 0.01§
groups groups groups groups groups
Legend: See text for explanation of graft regions. *Dunnett's t test not significant. tDunnett's t test: Group I vs. Group III = 5.76, p < 0.01. tDunnett's t test: Group I vs. Group III = 4.174, p < 0.05. §Dunnett's t test: Group I vs. Group III = 4.718, p < 0.05.
Table II. Intimal thickness: Graft region A (expressed in millimeters) Experimental group Group Group Group Group
I II III IV
No.
X
S.D.
S.E.
5 5 3 4
0.022 0.038 0.074 0.104
0.011 0.015 0.095 0.071
0.005 0.007 0.055 0.035
distal graft (D), and distal anastomosis (E). Tissue blocks for electron microscopy were postfixed in 2 per cent osmium tetroxide in a 0.1M P0 4 buffer for one hour, dehydrated through graded alcohols, and embedded in epoxy resin. Sections were cut at 1 p with a Porter-Blum MT-2B ultramicrotome and stained with toluidine blue for orientation. Thin sections were cut between 400 to 600 A, mounted on grids, and stained with a solution of saturated uranyl acetate and lead citrate. Grids were examined and photographed with a Philips Model 300 electron microscope. Two or more blocks were taken from each area of the graft, embedded in paraffin, and processed for light microscopic evaluation. Sections were cut at 5 /A and stained with hematoxylin and eosin, Alcian blue, colloidal iron, allochrome, Gomori's trichrome, and Weigert's resorcin-fuchsin. Frozen sections were taken from selective areas of the graft wall, cut at 5 p. and stained with oil red O for demonstration of lipids. Hematoxylin and eosin, Gomori's trichrome, and Weigert's resorcin-fuchsin stains were used to determine the limits of the tunica intima and media. The intimal-medial interface was identified on the basis of cellular orientation. Intimal thicknesses were measured from three areas of each trichrome-stained slide.
9 1 0 Bond et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 3. Light photomicrograph from the proximal graft (B) of a Group II animal showing a paucity of elastic fibers in the tunica intima. (Weigert's resorcin-fuchsin, x270.) Fig. 2. Histologic appearance of the midgraft (C) of a Group I animal showing a thin tunica intima (longitudinal fibers underlying the endothelium). (Hematoxylin and eosin, x270.) Measurements were made with an eyepiece micrometer calibrated with a known standard. Micrometer graduation intervals were determined for those combinations of occulars, tube length, and objectives used in this study. Alcian blue, colloidal iron, and allochromestained sections were used to determine the relative distribution and density of glycosaminoglycans. Results Gross changes in the vein graft. When the dogs were put to death, 6 months postoperatively, two Group III grafts and one Group IV graft were thrombosed and displayed evidence of recanalization. All patent grafts were markedly dilated when compared to the original diameter of the vein graft. Group I grafts showed an average increase in diameter of 52 per cent; Group II, 52 per cent; Group III, 49 per cent; and Group IV, 61 per cent.
Luminal surfaces overlying the proximal and distal anastomoses were similar in all grafts and consisted of an opaque, white, thickened intimal layer which extended for a distance of 2 mm. on either side of the suture lines. Endothelial surfaces of all grafts appeared similar. Few intact valves could be demonstrated grossly, but slightly raised, opaque areas similar in distribution to venous bicuspid valves were noted in all grafts. Light microscopy. Normal intima. The tunica intima of the normal femoral vein in the dog is composed of endothelium and a subendothelial space containing infrequent, obliquely oriented smooth muscle cells, collagen fibers, and small elastic fibers, most of which are arranged parallel to the longitudinal axis of the vessel. No clear demarcation between the tunica intima and media was observed. Graft intima. Hematoxylin and eosin and trichrome stains from the proximal and distal anastomoses (A and E, respectively) disclosed large amounts of collagen in
Volume 71 Number 6 June, 1976
Fig. 4 Increased elastic tissue of the tunica intima from the proximal graft (B) of a Group III animal. (Weigert's resorcin-fuchsin, X270.) the intima and media which extended to the level of the suture material. Frequently, there was a foreign body giant cell response associated with the sutures. The fibrous thickening extended on either side of the suture lines for a distance of not more than 2 mm. Statistical analysis of intimal thickness overlying the proximal anastomosis (A) failed to disclose significant differences among experimental groups (Tables I and II). Intimal thickness overlying the distal anastomosis (E) in Group III grafts, 0.156 ± 0.029 mm. (mean ± standard error), was significantly different (p < 0.01) from that of similar sites from all other groups and from Group I grafts, 0.030 ± 0.009, according to Dunnett's t test17 (p < 0.05) (Tables I and VI). Sections from the grafts proper in all groups contained a well-demarcated intimal layer characterized by longitudinally oriented smooth muscle cells with varying amounts of connective tissue fibers. Variability of intimal thickness was observed not only between groups (Table I) but within individual groups
Intimal proliferation in femoral bypass grafts
91 1
Fig. 5. Tunica intima of the distal graft (D) in a Group III animal showing increased cellularity. (Gomori's trichrome, X430.)
(Tables II to VI). Statistically different intimal thicknesses in proximal segments of grafts (B) were observed between all groups (analysis of variance, p < 0.01) and specifically between Group III (0.111 ± 0.025) and Group I (0.014 ± 0.003) (Tables I and III). Comparisons of intimal thickness between groups at midgraft levels (C) failed to demonstrate significant differences (Table IV). Thickness of the tunica intima within the distal graft (D) of Group HI animals, 0.157 ± 0.020, differed significantly from those of Group I, 0.051 ± 0.010 (one-way analysis of variance, p < 0.01, and Dunnett's " t " test, p < 0.05) (Tables I and V). Occasional sections from the graft wall (B, C, and D) in Groups I, II, and IV disclosed a very thin tunica intima composed of a single layer of endothelium with one or two layers of smooth muscle in the subendothelial space (Fig. 2). Similar findings were also noted in Group III animals at midgraft level (C).
9 1 2 Bond et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 6. Histologic appearance of the stump of a lost valve from the midgraft (C) of a Group IV animal showing the presence of afibromuscularcap overlying the base of the valve. (Gomori's trichrome, x70.) Composition of tunica intima. All sections of the grafts from the four groups of animals contained elastic fibers. However, in those areas where the tunica intima was thin, relatively fewer elastic fibers were observed (Fig. 3). Increased numbers of elastic fibers were generally noted in areas where the tunica intima was thickened. However, this phenomenon was not restricted to these areas (Fig. 4). Trichrome-stained sections were evaluated to determine ratio of cellular to intercellular components in the tunica intima. Group III grafts at sites B and D consistently contained a greater ratio of cellular to intercellular components when compared to the other groups (Fig. 5). Connective tissue stains were evaluated to determine the relative amounts of intimal glycosaminoglycans. Although the majority of grafts contained these connective tissue elements, the relative amounts could not be related to angle of implantation, presence of arterial stenosis in the left external iliac artery, presence of venous valves whether lost or intact, or relative thicknesses of the tunica intima. Microscopic sections processed and stained for intimal lipids were consistently negative. Valves. A total of six valves were found in the Group I grafts, four of which were reduced to moundlike structures and two of which were intact. Similarly, Group II grafts contained four valves, three of which were lost and one intact; Group III grafts contained two valves, both of which were intact; and Group IV grafts
Fig. 7. Electron micrograph of the intima of a Group III animal showing several well-differentiated smooth muscle cells. Note compactness of cells. Distal vein graft (D). (x 16,200.)
showed three valves, two of which were lost. Intimal surfaces overlying the base of lost valves were covered with caplike areas of subendothelial fibromuscular proliferation (Fig. 6). Marked variability was noted in the pre- and postvalve intimal thicknesses, but this variability was not significant among the four groups of animals. Electron microscopy. The endothelium was intact and continuous in all sections examined. Evidence of mural thrombus formation was not observed in the intact grafts. Subendothelial connective tissue varied in composition. Areas of thickened intima contained almost pure populations of well-differentiated smooth muscle cells characterized by a large component of myofilaments, dense bodies in the periphery of the cytoplasm, minimal amounts of endoplasmic reticulum, and numerous micropinocytotic vesicles (Fig. 7).
Volume 71 Number 6 June, 1976
Intimal proliferation in femoral bypass grafts
913
Fig. 8. Electron micrograph showing the proximal graft (B) of a Group II animal. Only one or two layers of smooth Muscle cells are seen. Infrequent fibroblasts were also observed in the intima (lower left). (x8,300.) Table III. Intimal thickness: Graft region B (expressed in millimeters) Experimental group Group I Group II Group III Group IV
No.
X
5 5 3 4
0.014 0.026 0.111 0.023
Table V. Intimal thickness: Graft region D (expressed in millimeters)
S.D.
S.E.
0.006 0.012 0.043 0.019
0.003 0.005 0.025 0.010
Group I Group II Group III Group IV
Group Group Group Group
I II III IV
No.
x
S.D.
S.E.
5 5 3 4
0.051 0.040 0.157 0.043
0.023 0.023 0.034 0.042
0.010 0.010 0.020 0.021
Table VI. Intimal thickness: Graft region E (expressed in millimeters)
Table IV. Intimal thickness: Graft region C (expressed in millimeters) Experimental group
Experimental group
No.
X
S.D.
S.E.
5 5 3 4
0.024 0.077 0.124 0.051
0.009 0.067 0.074 0.040
0.004 0.030 0.042 0.020
Experimental group Group I Group II Group III Group IV
No.
X
S.D.
S.E.
5 5 3 4
0.030 0.051 0.156 0.078
0.020 0.034 0.050 0.031
0.009 0.015 0.029 0.051
Discussion muscle cells in addition to mature fibroblasts. However, the former were always the predominant cell type (Fig. 8). The tunica media was characterized by relatively large amounts of collagen when compared to the tunica intima. The predominant cell type in the media was the fibroblast, but a few smooth muscle cells were also observed (Fig. 9).
Venous bypass grafts failing late postoperatively have consistently demonstrated a thickened subendothelial fibromuscular layer.3- 5 - 7 - 9 - 10 - 15 - 16 - 18 ~ 21 - 23 25, 27 rphis layer developed in the experimental animals in our study, but only in the acute angle grafts (Group III) was the intimal thickness statistically different from that in the other groups.
9 1 4 Bond et al.
Fig. 9. Electron micrograph of the tunica media of a Group IV animal. Distal graft (D) showing fibroblasts with numerous collagen fibers. (x5,500.) The cause(s) for intimal thickening in arterialized veins is not known. Brody and associates23 have experimentally produced subendothelial fibromuscular thickening in cephalic and femoral veins of dogs by devascularizing the vein wall in situ. Interruption of vasa vasora with subsequent ischemic change in the vessel wall has been suggested by others as a possible etiology of intimal thickening.22 We suggest that factors in addition to interruption of vasa vasora are involved in the intimal proliferation observed in Group III grafts. The fact that all animals in Group III developed a significantly thicker intima in selective areas of the graft than did Group I, II, or IV animals suggests etiological factors which are associated with or result from the angle of implantation. Kennedy and associates16 have reported on 8 patients reoperated upon an average of 10.6 months after receiving aorta-coronary bypass grafts. The average angle at the proximal anastomosis in the initial grafts was 45.8
The Journal of Thoracic and Cardiovascular Surgery
degrees and the average occlusion of the grafts was 84.4 per cent as demonstrated by angiography. These same authors have studied the optimal angle of anastomosis as predicted by (1) an in vitro study with a two-dimensional flow model and (2) a computer simulation that matches coronary resistance to aortic pressure with varying angles of anastomosis. They suggest that the optimal theoretical angle at the proximal anastomosis is 90 degrees. This study was not designed to elucidate hemodynamic stresses occurring in vein grafts. However, low flow velocity and increased wall stress within arteriovenous grafts may be significant factors in the development of fibromuscular intimal thickening.24 Continuation of intimal lesions from anastomosis into adjacent arteries has previously been described.26, 27 In the present study these lesions were not impressive in terms of size and were composed primarily of collagen with relatively few smooth muscle cells. Significant differences were not observed in the thickness or composition of these arterial lesions among the experimental groups. The mild intimal thickening found in Group II dogs was unexpected. We had hypothesized at the outset of these experiments that runoff obstruction would stimulate intimal thickening. Obstruction had very little or no effect on the magnitude of intimal thickening. Our findings compare well with clinical studies which did not relate late failure of grafts with poor runoff vessels.1 We were unable to document significant changes in the pre- and postvalve regions of the grafts either among or within experimental groups. Although focal areas of fibromuscular hyperplasia were occasionally noted associated with intact valves, the relatively few animals having these lesions precludes a definitive statement. Valves which were spontaneously lost during the course of the experiment were replaced by a caplike area of intimal fibromuscular hyperplasia. These lesions, in addition to the anastomotic sites, were the only areas of the graft wall that protruded into the lumen. It is tempting to speculate that concentric, segmental stenosis in arteriovenous grafts results from these lesions. Our studies suggest that focal endothelial trauma and interruption of vasa vasora following surgical transplantation may play a role in the development of diffuse intimal fibromuscular hyperplasia in autologous vein grafts; however, they further indicate that these factors alone are not responsible for the severe, diffuse,
Volume 71 Number 6 June, 1976
and often segmental stenosis frequently found in bypass grafts. We suggest that segmental distribution of intimal fibromuscular proliferation found in Group III grafts is, at least in part, the result of hemodynamic stresses on selective regions of the graft wall resulting from angles of less than 90 degrees at the proximal anastomosis.
Intimal proliferation
9
10
Summary Autologous vein bypasses between external iliac arteries of dogs were studied to determine effects of angulation and runoff stenosis on the development of intimal thickening in the grafts. The magnitude of intimal thickening was small for both variables and significantly greater only in grafts placed at an acute angle. The proximal and distal graft segments as well as the distal anastomosis in the acute angle bypasses are selective sites of increased intimal thickening. We wish to express our appreciation to Ms. Barbara Wheaton for processing the necessary histological material, to Dr. Jean D. Hensel, Assistant Professor of Biostatistics, Ohio State University, for her statistical analysis, and to Norma Lofland and Shirley Pegram for preparing the manuscript. REFERENCES 1 Lesperance, J., Bourassa, M. G., Biron, P., Campeau, L., and Saltiel, J.: Aorta to Coronary Artery Saphenous Vein Grafts: Preoperative Angiographic Criteria for Successful Surgery, Am. J. Cardiol. 30: 459, 1972. 2 Hamaker, W. R., Doyle, W. F., O'Connell, T. J., and Gomez, A. C : Subintimal Obliterative Proliferation in Saphenous Vein Grafts, Ann. Thorac. Surg. 13: 488, 1972. 3 Vlodaver, Z., and Edwards, J. E.: Pathologic Changes in Aortic-Coronary Arterial Saphenous Vein Grafts, Circulation 44: 719, 1971. 4 Sheldon, W. C , Rincon, G., Effler, D. B., Proudfit, W. L., and Sones, F. M.: Vein Graft Surgery for Coronary Artery Disease: Survival and Angiographic Results in 1,000 Patients, Circulation 48: 184, 1973 (Suppl. III). 5 Flemma, R. J., Johnson, W. D., Lepley, D., Tector, A. J., Walker, J., Gale, H., Beddingfield, G., and Manley, J. C : Saphenous Vein Bypass Grafting for Myocardial Revascularization, Ann. Thorac. Surg. 14: 232, 1973. 6 Grondin, C M . , Castonguay, Y. R., Lepage, G., Meere, C., and Grondin, P.: Aortocoronary Bypass Grafts, Arch. Surg. 103: 535, 1971. 7 Johnson, W. D., Aver, J. E., and Tector, A. J.: Late Changes in Coronary Vein Grafts, Am. J. Cardiol. 26: 640, 1970. 8 Adam, M., Geisler, G. F., Lambert, C. J., and Mitchell,
11 12
13
14 15
16
in femoral bypass grafts
9 15
B. F.: Reoperation Following Clinical Failure of Aortoto-Coronary Artery Bypass Vein Grafts, Ann. Thorac. Surg. 14: 272, 1972. Grondin, C. M., Meere, C , Castonguay, Y., Lepage, G., and Grondin, P.: Progressive and Late Obstruction of an Aorto-Coronary Venous Bypass Graft, Circulation 43: 698, 1971. Kern, W. H., Dermer, G. B., and Lindesmith, G. G.: The Intimal Proliferation in Aortic Coronary Saphenous Vein Grafts: Light and Electron Microscopic Studies, Am. Heart J. 84: 771, 1972. Linton, R. R., and Darling, R. C : Autogenous Saphenous Vein Bypass Grafts in Femoropopliteal Obliterative Arterial Disease, Surgery 51: 62, 1962. Baddeley, R. M., Ashton, F., Slaney, G., and Barnes, A. D.: Late Results of Autogenous Vein Bypass Grafts in Femoropopliteal Arterial Occlusion, Br. Med. J. 1: 653, 1970. Baddeley, R. M., Lawson, L. J., Ashton, F., and Slaney, G.: Evaluation of Autogenous Vein Bypass Grafts for Femoro-popliteal Arterial Occlusion, Br. Med. J. 2: 410, 1967. Hall, K. V.: The Great Saphenous Vein Used In Situ as an Arterial Shunt After Vein Valve Extirpation: A Follow-up Study, Acta Chir. Scand. 129: 33, 1965. McNamara, J. J., Darling, R. C , and Linton, R. R.: Segmental Stenosis of Saphenous-Vein Autografts: Preventable Cause of Late Occlusion in Arterial Reconstruction, N. Engl. J. Med. 277: 290, 1967. Kennedy, J. H., Wieting, D. W., Hwang, N. H. C , Anderson, M. S., Bayardo, R. J., Howell, J. F., and De Bakey, M. E.: Hydraulic and Morphologic Study of Fibrous Intimal Hyperplasia in Autogenous Saphenous Vein Bypass Grafts, J. THORAC. CARDIOVASC. SURG. 67:
17 18 19
20 21
22 23
805, 1974. Winer, B. J.: Statistical Principles in Experimental Design, New York, 1962, McGraw-Hill Book Company, Inc., pp. 201-205. Bond, M. G.: Vascular Changes Related to Angle of Implantation and Arterial Stenosis in Arterial Venous Grafts, Anat. Rec. 178: 314a, 1974. Bond, M. G., Karayannacos, P., Hostetler, J., Geer, J., and Vasko, J.: Histological Changes Related to Degree of Angulation and Distal Stenosis in Arterial Venous Grafts, Clin. Res. 22: 265a, 1974. Brody, W. R., Angell, W. W., and Kosek, J. C : Histologic Fate of the Venous Coronary Artery Bypass in Dogs, Am. J. Pathol. 66: 111, 1972. McCabe, M., Cunningham, G. J., Wyatt, A. P., Rothnie, N. G., and Taylor, G. W.: A Histological and Histochemical Examination of Autogenous Vein Grafts, Br. J. Surg. 54: 147, 1967. Jones, T. I., and Dale, W. A.: Study of Peripheral Autogenous Vein Grafts, Arch. Surg. 76: 294, 1958. Brody, W. R., Kosek, J. C , and Angell, W. W.: Changes in Vein Grafts Following Aorto-coronary
9 16
The Journal of Thoracic and Cardiovascular Surgery
Bond et al.
Bypass Induced by Pressure and Ischemia, J. THORAC. CARDIOVASC. SURG. 64: 847,
1972.
24 Karayannacos, P. E., Geer, J., Gast, M., Hodges, R., Bond, G., and Vasko, J. S.: Wall Strain in Arterial Vein Grafts, Clin. Res. 21: 813, 1973. 25 Conkle, D. M., Jones, M., Levin, F. H., Melvin, D. B., Stinson, E. B., Ferans, V. J., and Roberts, W.C.: Subendothelial Lesions Observed in Arterial Autogenous Vein Grafts: Light and Electron Microscopic Evaluation, Circulation 46: 4, 1972 (Suppl. II).
26 Jones, M., Conkle, D. M., Ferrans, V. J., Roberts, W. C , Levine, F. H., Melvin, D. B., and Stinson, E. B.: Lesions Observed in Arterial Autogenous Vein Grafts, Circulation 48: 198, 1973 (Suppl. III). 27 Lesperance, J., Bourassa, M. G., Saltiel, J., and Grondin, C. M.: Late Changes in Aorto-Coronary Vein Grafts: Angiographic Features, Am. J. Roentgenol. Radium Ther. Nucl. Med. 116: 74, 1972.
M. G. Bond, Ph.D.,* J. R. Hostetler, Ph.D.,* P. E. Karayannacos, M.D.,** J. C. Geer, M.D.,*** and J. S. Vasko, M.D., ** Columbus, Ohio
Zxutologus vein bypass surgery for the treatment of obliterative arterial disease involving coronary and peripheral arteries is being performed with increasing frequency. Late failure of venous grafts has been reported in aorta-coronary bypasses1-10 and arterial bypasses of the lower extremities.11-14 These vein grafts often show the presence of a subendothelial fibromuscular layer which may effectively reduce lumen diameter. Various etiological factors suggested to explain this phenomenon have included angulation of the proximal anastomosis,15'16 turbulence in the area
of intact valves,2 trauma to the vein graft during the surgical procedure,15 and the presence of obstructive lesions in the arterial outflow tract.3 The purpose of this investigation is to determine what changes occur in the thickness and distribution of the subendothelial fibromuscular layer that result from (1) different angles of implantation at the proximal anastomosis, (2) stenosis in the distal runoff artery, and (3) the presence of valves within the reversed vein graft.
From the Departments of Anatomy, Pathology, and Surgery, The Ohio State University, Columbus, Ohio 43210. Supported in part by National Heart and Lung Institute Grant HL11897, Harker's Fund Grant, and Central Ohio Heart Chapter Grant 3554-A1.
Twenty male and female mongrel dogs, weighing between 20 and 25 kilograms, received autologous femoral vein bypass grafts between the external iliac arteries. Animals were equally divided into four groups based upon the angulation of the proximal anastomosis and the presence of 80 per cent stenosis in the runoff artery (Fig. 1). Group I animals received grafts at a 90 degree angle relative to the proximal artery, Group II animals received 90 degree angle grafts besides being subjected to an 80 per cent stenosis in the runoff artery, Group III animals received grafts at an acute angle (55 degrees ± 10 degrees), and Group IV animals re-
Received for publication Dec. 24, 1975. Accepted for publication Jan. 29, 1976. Address for reprints: M. G. Bond, Ph.D., Department of Comparative Medicine, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, N. C. 27103. *Department of Anatomy. "Division of Thoracic Surgery, Department of Surgery. ***Department of Pathology.
Methods
907
The Journal of Thoracic and Cardiovascular Surgery
9 0 8 Bond et al.
Aorta Right external iliac artery
Stenosis AV Graft Group IT, 90°angle+ 8 0 % stenosis
Group I , 90° angle
Group M, 55° angle Group W,
Fig. 1. Diagram of experimental model. AV, Arteriovenous.
ceived grafts at an obtuse angle (120 degrees ± 10). Thirty minutes before the operation the animals were premedicated with acepromazine (0.55 mg. per kilogram) and atropine (0.05 mg. per kilogram). Anesthesia was initiated with sodium thiopental (20 mg. per kilogram). An endotracheal tube was placed and connected to a volume-controlled respirator. Anesthesia was maintained with a mixture of nitrous oxide, halothane, and oxygen. A low midline laparotomy was performed under sterile conditions and the retroperitoneal space was entered distal to the aortic bifurcation. The external iliac arteries were mobilized and umbilical tapes placed around them. A skin incision was made over the right inguinocrural region and the femoral vein identified and mobilized over a length of 10 cm. The vein was dissected and handled with great care to avoid unnecessary injury to the vessel wall. Tributaries were ligated with 6-0 polypropylene monofilament suture material. The section of femoral vein was removed and flushed with heparinized saline, with special care taken to avoid overdistention. The vein segment was placed in saline at room temperature
until the femoral arteries were prepared for anastomosis. A segment of right external iliac artery was isolated between two atraumatic vascular clamps and a longitudinal arteriotomy was performed on the medial aspect of the artery. An end-to-side anastomosis was accomplished between the iliac artery and the distal end of the vein graft by use of a continuous suture with 6-0 polypropylene monofilament. At the completion of the anastomosis, the graft was cross-clamped near the proximal end and flow was restored through the right external iliac artery. The graft was passed through a surgical opening in the mesorectum and brought into apposition with the left external iliac artery, where a similar vascular anastomosis was performed. All clamps were removed and the left external iliac artery was twice ligated with 2-0 silk proximal to the anastomotic site. Blood flow was now from the right external iliac artery, designated the proximal artery, through the graft to the left external iliac artery, now designated the distal artery. The mesorectum was reconstructed with interrupted
Volume 71 Number 6 June, 1976
Intimal proliferation in femoral bypass grafts
3-0 catgut stitches. The peritoneum was closed with continuous 2-0 chromic catgut and the linea alba apposed with interrupted 3-0 Dacron. Skin was closed with Vetafil, a special veterinary suture material. Elapsed time between removal of the vein and completion of the graft anastomosis was approximately 30 minutes. Distances of the anastomotic sites from the aortic bifurcation and shapes of the proximal and distal ends of the vein graft varied depending on the angle between the proximal artery and the graft. The animals received antibiotic therapy with penicillin and streptomycin for one week, and sutures were removed on the tenth postoperative day. Ten days postoperatively, 5 animals with 90 degree angle anastomoses were reoperated upon. Under the same anesthetic and sterile surgical conditions, a low left paramedian incision was made. The left distal external iliac artery was identified and mobilized. A 60 per cent constriction in diameter of the artery, corresponding to an 80 per cent reduction in crosssectional area of the vessel, was achieved by a circumferential Mersilene tape 5 mm. wide. The abdomen was then closed in the described manner. All animals received postoperative treatment similar to that after the initial operation. Six months from the time of the original operation animals from all groups were again operated upon, under sodium nembutal anesthesia. Through a midline laparotomy the abdominal aorta, external iliac arteries, and the vein graft were dissected. The abdominal aorta was cross-clamped and a large cannula inserted into the vessel distal to the clamp. The posterior extremities and pelvis were perfused in situ under a pressure of 100 mm. Hg with 4° C. normal saline followed by a 3 per cent glutaraldehyde-formaldehyde fixing solution. Specimens were washed in 0.1 M P0 4 buffer solution (pH 7.4) at room temperature for 24 to 48 hours. The diameter of the graft was measured at operation with a caliper and the true diameter calculated after subtracting the thickness of the vein wall. After fixation, the graft was removed and opened longitudinally. Internal circumference was measured with a micrometer. The diameter was derived from the formula: circumference =
2TTR.
A 25 per cent shrinkage from the fixation process was considered. Representative sections for light and electron microscopic evaluation were taken from the proximal anastomosis (A), proximal graft (B), midgraft (C),
909
Table I. One-way analysis of variance for intimal thickening Region of graft A B C D E
Source of variation Between Between Between Between Between
Degrees of freedom
F
Level of significance
3 3 3 3 3
2.206 15.977 2.595 11.428 10.076
NS* p < O.Olt NS* p < 0.01$ p < 0.01§
groups groups groups groups groups
Legend: See text for explanation of graft regions. *Dunnett's t test not significant. tDunnett's t test: Group I vs. Group III = 5.76, p < 0.01. tDunnett's t test: Group I vs. Group III = 4.174, p < 0.05. §Dunnett's t test: Group I vs. Group III = 4.718, p < 0.05.
Table II. Intimal thickness: Graft region A (expressed in millimeters) Experimental group Group Group Group Group
I II III IV
No.
X
S.D.
S.E.
5 5 3 4
0.022 0.038 0.074 0.104
0.011 0.015 0.095 0.071
0.005 0.007 0.055 0.035
distal graft (D), and distal anastomosis (E). Tissue blocks for electron microscopy were postfixed in 2 per cent osmium tetroxide in a 0.1M P0 4 buffer for one hour, dehydrated through graded alcohols, and embedded in epoxy resin. Sections were cut at 1 p with a Porter-Blum MT-2B ultramicrotome and stained with toluidine blue for orientation. Thin sections were cut between 400 to 600 A, mounted on grids, and stained with a solution of saturated uranyl acetate and lead citrate. Grids were examined and photographed with a Philips Model 300 electron microscope. Two or more blocks were taken from each area of the graft, embedded in paraffin, and processed for light microscopic evaluation. Sections were cut at 5 /A and stained with hematoxylin and eosin, Alcian blue, colloidal iron, allochrome, Gomori's trichrome, and Weigert's resorcin-fuchsin. Frozen sections were taken from selective areas of the graft wall, cut at 5 p. and stained with oil red O for demonstration of lipids. Hematoxylin and eosin, Gomori's trichrome, and Weigert's resorcin-fuchsin stains were used to determine the limits of the tunica intima and media. The intimal-medial interface was identified on the basis of cellular orientation. Intimal thicknesses were measured from three areas of each trichrome-stained slide.
9 1 0 Bond et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 3. Light photomicrograph from the proximal graft (B) of a Group II animal showing a paucity of elastic fibers in the tunica intima. (Weigert's resorcin-fuchsin, x270.) Fig. 2. Histologic appearance of the midgraft (C) of a Group I animal showing a thin tunica intima (longitudinal fibers underlying the endothelium). (Hematoxylin and eosin, x270.) Measurements were made with an eyepiece micrometer calibrated with a known standard. Micrometer graduation intervals were determined for those combinations of occulars, tube length, and objectives used in this study. Alcian blue, colloidal iron, and allochromestained sections were used to determine the relative distribution and density of glycosaminoglycans. Results Gross changes in the vein graft. When the dogs were put to death, 6 months postoperatively, two Group III grafts and one Group IV graft were thrombosed and displayed evidence of recanalization. All patent grafts were markedly dilated when compared to the original diameter of the vein graft. Group I grafts showed an average increase in diameter of 52 per cent; Group II, 52 per cent; Group III, 49 per cent; and Group IV, 61 per cent.
Luminal surfaces overlying the proximal and distal anastomoses were similar in all grafts and consisted of an opaque, white, thickened intimal layer which extended for a distance of 2 mm. on either side of the suture lines. Endothelial surfaces of all grafts appeared similar. Few intact valves could be demonstrated grossly, but slightly raised, opaque areas similar in distribution to venous bicuspid valves were noted in all grafts. Light microscopy. Normal intima. The tunica intima of the normal femoral vein in the dog is composed of endothelium and a subendothelial space containing infrequent, obliquely oriented smooth muscle cells, collagen fibers, and small elastic fibers, most of which are arranged parallel to the longitudinal axis of the vessel. No clear demarcation between the tunica intima and media was observed. Graft intima. Hematoxylin and eosin and trichrome stains from the proximal and distal anastomoses (A and E, respectively) disclosed large amounts of collagen in
Volume 71 Number 6 June, 1976
Fig. 4 Increased elastic tissue of the tunica intima from the proximal graft (B) of a Group III animal. (Weigert's resorcin-fuchsin, X270.) the intima and media which extended to the level of the suture material. Frequently, there was a foreign body giant cell response associated with the sutures. The fibrous thickening extended on either side of the suture lines for a distance of not more than 2 mm. Statistical analysis of intimal thickness overlying the proximal anastomosis (A) failed to disclose significant differences among experimental groups (Tables I and II). Intimal thickness overlying the distal anastomosis (E) in Group III grafts, 0.156 ± 0.029 mm. (mean ± standard error), was significantly different (p < 0.01) from that of similar sites from all other groups and from Group I grafts, 0.030 ± 0.009, according to Dunnett's t test17 (p < 0.05) (Tables I and VI). Sections from the grafts proper in all groups contained a well-demarcated intimal layer characterized by longitudinally oriented smooth muscle cells with varying amounts of connective tissue fibers. Variability of intimal thickness was observed not only between groups (Table I) but within individual groups
Intimal proliferation in femoral bypass grafts
91 1
Fig. 5. Tunica intima of the distal graft (D) in a Group III animal showing increased cellularity. (Gomori's trichrome, X430.)
(Tables II to VI). Statistically different intimal thicknesses in proximal segments of grafts (B) were observed between all groups (analysis of variance, p < 0.01) and specifically between Group III (0.111 ± 0.025) and Group I (0.014 ± 0.003) (Tables I and III). Comparisons of intimal thickness between groups at midgraft levels (C) failed to demonstrate significant differences (Table IV). Thickness of the tunica intima within the distal graft (D) of Group HI animals, 0.157 ± 0.020, differed significantly from those of Group I, 0.051 ± 0.010 (one-way analysis of variance, p < 0.01, and Dunnett's " t " test, p < 0.05) (Tables I and V). Occasional sections from the graft wall (B, C, and D) in Groups I, II, and IV disclosed a very thin tunica intima composed of a single layer of endothelium with one or two layers of smooth muscle in the subendothelial space (Fig. 2). Similar findings were also noted in Group III animals at midgraft level (C).
9 1 2 Bond et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 6. Histologic appearance of the stump of a lost valve from the midgraft (C) of a Group IV animal showing the presence of afibromuscularcap overlying the base of the valve. (Gomori's trichrome, x70.) Composition of tunica intima. All sections of the grafts from the four groups of animals contained elastic fibers. However, in those areas where the tunica intima was thin, relatively fewer elastic fibers were observed (Fig. 3). Increased numbers of elastic fibers were generally noted in areas where the tunica intima was thickened. However, this phenomenon was not restricted to these areas (Fig. 4). Trichrome-stained sections were evaluated to determine ratio of cellular to intercellular components in the tunica intima. Group III grafts at sites B and D consistently contained a greater ratio of cellular to intercellular components when compared to the other groups (Fig. 5). Connective tissue stains were evaluated to determine the relative amounts of intimal glycosaminoglycans. Although the majority of grafts contained these connective tissue elements, the relative amounts could not be related to angle of implantation, presence of arterial stenosis in the left external iliac artery, presence of venous valves whether lost or intact, or relative thicknesses of the tunica intima. Microscopic sections processed and stained for intimal lipids were consistently negative. Valves. A total of six valves were found in the Group I grafts, four of which were reduced to moundlike structures and two of which were intact. Similarly, Group II grafts contained four valves, three of which were lost and one intact; Group III grafts contained two valves, both of which were intact; and Group IV grafts
Fig. 7. Electron micrograph of the intima of a Group III animal showing several well-differentiated smooth muscle cells. Note compactness of cells. Distal vein graft (D). (x 16,200.)
showed three valves, two of which were lost. Intimal surfaces overlying the base of lost valves were covered with caplike areas of subendothelial fibromuscular proliferation (Fig. 6). Marked variability was noted in the pre- and postvalve intimal thicknesses, but this variability was not significant among the four groups of animals. Electron microscopy. The endothelium was intact and continuous in all sections examined. Evidence of mural thrombus formation was not observed in the intact grafts. Subendothelial connective tissue varied in composition. Areas of thickened intima contained almost pure populations of well-differentiated smooth muscle cells characterized by a large component of myofilaments, dense bodies in the periphery of the cytoplasm, minimal amounts of endoplasmic reticulum, and numerous micropinocytotic vesicles (Fig. 7).
Volume 71 Number 6 June, 1976
Intimal proliferation in femoral bypass grafts
913
Fig. 8. Electron micrograph showing the proximal graft (B) of a Group II animal. Only one or two layers of smooth Muscle cells are seen. Infrequent fibroblasts were also observed in the intima (lower left). (x8,300.) Table III. Intimal thickness: Graft region B (expressed in millimeters) Experimental group Group I Group II Group III Group IV
No.
X
5 5 3 4
0.014 0.026 0.111 0.023
Table V. Intimal thickness: Graft region D (expressed in millimeters)
S.D.
S.E.
0.006 0.012 0.043 0.019
0.003 0.005 0.025 0.010
Group I Group II Group III Group IV
Group Group Group Group
I II III IV
No.
x
S.D.
S.E.
5 5 3 4
0.051 0.040 0.157 0.043
0.023 0.023 0.034 0.042
0.010 0.010 0.020 0.021
Table VI. Intimal thickness: Graft region E (expressed in millimeters)
Table IV. Intimal thickness: Graft region C (expressed in millimeters) Experimental group
Experimental group
No.
X
S.D.
S.E.
5 5 3 4
0.024 0.077 0.124 0.051
0.009 0.067 0.074 0.040
0.004 0.030 0.042 0.020
Experimental group Group I Group II Group III Group IV
No.
X
S.D.
S.E.
5 5 3 4
0.030 0.051 0.156 0.078
0.020 0.034 0.050 0.031
0.009 0.015 0.029 0.051
Discussion muscle cells in addition to mature fibroblasts. However, the former were always the predominant cell type (Fig. 8). The tunica media was characterized by relatively large amounts of collagen when compared to the tunica intima. The predominant cell type in the media was the fibroblast, but a few smooth muscle cells were also observed (Fig. 9).
Venous bypass grafts failing late postoperatively have consistently demonstrated a thickened subendothelial fibromuscular layer.3- 5 - 7 - 9 - 10 - 15 - 16 - 18 ~ 21 - 23 25, 27 rphis layer developed in the experimental animals in our study, but only in the acute angle grafts (Group III) was the intimal thickness statistically different from that in the other groups.
9 1 4 Bond et al.
Fig. 9. Electron micrograph of the tunica media of a Group IV animal. Distal graft (D) showing fibroblasts with numerous collagen fibers. (x5,500.) The cause(s) for intimal thickening in arterialized veins is not known. Brody and associates23 have experimentally produced subendothelial fibromuscular thickening in cephalic and femoral veins of dogs by devascularizing the vein wall in situ. Interruption of vasa vasora with subsequent ischemic change in the vessel wall has been suggested by others as a possible etiology of intimal thickening.22 We suggest that factors in addition to interruption of vasa vasora are involved in the intimal proliferation observed in Group III grafts. The fact that all animals in Group III developed a significantly thicker intima in selective areas of the graft than did Group I, II, or IV animals suggests etiological factors which are associated with or result from the angle of implantation. Kennedy and associates16 have reported on 8 patients reoperated upon an average of 10.6 months after receiving aorta-coronary bypass grafts. The average angle at the proximal anastomosis in the initial grafts was 45.8
The Journal of Thoracic and Cardiovascular Surgery
degrees and the average occlusion of the grafts was 84.4 per cent as demonstrated by angiography. These same authors have studied the optimal angle of anastomosis as predicted by (1) an in vitro study with a two-dimensional flow model and (2) a computer simulation that matches coronary resistance to aortic pressure with varying angles of anastomosis. They suggest that the optimal theoretical angle at the proximal anastomosis is 90 degrees. This study was not designed to elucidate hemodynamic stresses occurring in vein grafts. However, low flow velocity and increased wall stress within arteriovenous grafts may be significant factors in the development of fibromuscular intimal thickening.24 Continuation of intimal lesions from anastomosis into adjacent arteries has previously been described.26, 27 In the present study these lesions were not impressive in terms of size and were composed primarily of collagen with relatively few smooth muscle cells. Significant differences were not observed in the thickness or composition of these arterial lesions among the experimental groups. The mild intimal thickening found in Group II dogs was unexpected. We had hypothesized at the outset of these experiments that runoff obstruction would stimulate intimal thickening. Obstruction had very little or no effect on the magnitude of intimal thickening. Our findings compare well with clinical studies which did not relate late failure of grafts with poor runoff vessels.1 We were unable to document significant changes in the pre- and postvalve regions of the grafts either among or within experimental groups. Although focal areas of fibromuscular hyperplasia were occasionally noted associated with intact valves, the relatively few animals having these lesions precludes a definitive statement. Valves which were spontaneously lost during the course of the experiment were replaced by a caplike area of intimal fibromuscular hyperplasia. These lesions, in addition to the anastomotic sites, were the only areas of the graft wall that protruded into the lumen. It is tempting to speculate that concentric, segmental stenosis in arteriovenous grafts results from these lesions. Our studies suggest that focal endothelial trauma and interruption of vasa vasora following surgical transplantation may play a role in the development of diffuse intimal fibromuscular hyperplasia in autologous vein grafts; however, they further indicate that these factors alone are not responsible for the severe, diffuse,
Volume 71 Number 6 June, 1976
and often segmental stenosis frequently found in bypass grafts. We suggest that segmental distribution of intimal fibromuscular proliferation found in Group III grafts is, at least in part, the result of hemodynamic stresses on selective regions of the graft wall resulting from angles of less than 90 degrees at the proximal anastomosis.
Intimal proliferation
9
10
Summary Autologous vein bypasses between external iliac arteries of dogs were studied to determine effects of angulation and runoff stenosis on the development of intimal thickening in the grafts. The magnitude of intimal thickening was small for both variables and significantly greater only in grafts placed at an acute angle. The proximal and distal graft segments as well as the distal anastomosis in the acute angle bypasses are selective sites of increased intimal thickening. We wish to express our appreciation to Ms. Barbara Wheaton for processing the necessary histological material, to Dr. Jean D. Hensel, Assistant Professor of Biostatistics, Ohio State University, for her statistical analysis, and to Norma Lofland and Shirley Pegram for preparing the manuscript. REFERENCES 1 Lesperance, J., Bourassa, M. G., Biron, P., Campeau, L., and Saltiel, J.: Aorta to Coronary Artery Saphenous Vein Grafts: Preoperative Angiographic Criteria for Successful Surgery, Am. J. Cardiol. 30: 459, 1972. 2 Hamaker, W. R., Doyle, W. F., O'Connell, T. J., and Gomez, A. C : Subintimal Obliterative Proliferation in Saphenous Vein Grafts, Ann. Thorac. Surg. 13: 488, 1972. 3 Vlodaver, Z., and Edwards, J. E.: Pathologic Changes in Aortic-Coronary Arterial Saphenous Vein Grafts, Circulation 44: 719, 1971. 4 Sheldon, W. C , Rincon, G., Effler, D. B., Proudfit, W. L., and Sones, F. M.: Vein Graft Surgery for Coronary Artery Disease: Survival and Angiographic Results in 1,000 Patients, Circulation 48: 184, 1973 (Suppl. III). 5 Flemma, R. J., Johnson, W. D., Lepley, D., Tector, A. J., Walker, J., Gale, H., Beddingfield, G., and Manley, J. C : Saphenous Vein Bypass Grafting for Myocardial Revascularization, Ann. Thorac. Surg. 14: 232, 1973. 6 Grondin, C M . , Castonguay, Y. R., Lepage, G., Meere, C., and Grondin, P.: Aortocoronary Bypass Grafts, Arch. Surg. 103: 535, 1971. 7 Johnson, W. D., Aver, J. E., and Tector, A. J.: Late Changes in Coronary Vein Grafts, Am. J. Cardiol. 26: 640, 1970. 8 Adam, M., Geisler, G. F., Lambert, C. J., and Mitchell,
11 12
13
14 15
16
in femoral bypass grafts
9 15
B. F.: Reoperation Following Clinical Failure of Aortoto-Coronary Artery Bypass Vein Grafts, Ann. Thorac. Surg. 14: 272, 1972. Grondin, C. M., Meere, C , Castonguay, Y., Lepage, G., and Grondin, P.: Progressive and Late Obstruction of an Aorto-Coronary Venous Bypass Graft, Circulation 43: 698, 1971. Kern, W. H., Dermer, G. B., and Lindesmith, G. G.: The Intimal Proliferation in Aortic Coronary Saphenous Vein Grafts: Light and Electron Microscopic Studies, Am. Heart J. 84: 771, 1972. Linton, R. R., and Darling, R. C : Autogenous Saphenous Vein Bypass Grafts in Femoropopliteal Obliterative Arterial Disease, Surgery 51: 62, 1962. Baddeley, R. M., Ashton, F., Slaney, G., and Barnes, A. D.: Late Results of Autogenous Vein Bypass Grafts in Femoropopliteal Arterial Occlusion, Br. Med. J. 1: 653, 1970. Baddeley, R. M., Lawson, L. J., Ashton, F., and Slaney, G.: Evaluation of Autogenous Vein Bypass Grafts for Femoro-popliteal Arterial Occlusion, Br. Med. J. 2: 410, 1967. Hall, K. V.: The Great Saphenous Vein Used In Situ as an Arterial Shunt After Vein Valve Extirpation: A Follow-up Study, Acta Chir. Scand. 129: 33, 1965. McNamara, J. J., Darling, R. C , and Linton, R. R.: Segmental Stenosis of Saphenous-Vein Autografts: Preventable Cause of Late Occlusion in Arterial Reconstruction, N. Engl. J. Med. 277: 290, 1967. Kennedy, J. H., Wieting, D. W., Hwang, N. H. C , Anderson, M. S., Bayardo, R. J., Howell, J. F., and De Bakey, M. E.: Hydraulic and Morphologic Study of Fibrous Intimal Hyperplasia in Autogenous Saphenous Vein Bypass Grafts, J. THORAC. CARDIOVASC. SURG. 67:
17 18 19
20 21
22 23
805, 1974. Winer, B. J.: Statistical Principles in Experimental Design, New York, 1962, McGraw-Hill Book Company, Inc., pp. 201-205. Bond, M. G.: Vascular Changes Related to Angle of Implantation and Arterial Stenosis in Arterial Venous Grafts, Anat. Rec. 178: 314a, 1974. Bond, M. G., Karayannacos, P., Hostetler, J., Geer, J., and Vasko, J.: Histological Changes Related to Degree of Angulation and Distal Stenosis in Arterial Venous Grafts, Clin. Res. 22: 265a, 1974. Brody, W. R., Angell, W. W., and Kosek, J. C : Histologic Fate of the Venous Coronary Artery Bypass in Dogs, Am. J. Pathol. 66: 111, 1972. McCabe, M., Cunningham, G. J., Wyatt, A. P., Rothnie, N. G., and Taylor, G. W.: A Histological and Histochemical Examination of Autogenous Vein Grafts, Br. J. Surg. 54: 147, 1967. Jones, T. I., and Dale, W. A.: Study of Peripheral Autogenous Vein Grafts, Arch. Surg. 76: 294, 1958. Brody, W. R., Kosek, J. C , and Angell, W. W.: Changes in Vein Grafts Following Aorto-coronary
9 16
The Journal of Thoracic and Cardiovascular Surgery
Bond et al.
Bypass Induced by Pressure and Ischemia, J. THORAC. CARDIOVASC. SURG. 64: 847,
1972.
24 Karayannacos, P. E., Geer, J., Gast, M., Hodges, R., Bond, G., and Vasko, J. S.: Wall Strain in Arterial Vein Grafts, Clin. Res. 21: 813, 1973. 25 Conkle, D. M., Jones, M., Levin, F. H., Melvin, D. B., Stinson, E. B., Ferans, V. J., and Roberts, W.C.: Subendothelial Lesions Observed in Arterial Autogenous Vein Grafts: Light and Electron Microscopic Evaluation, Circulation 46: 4, 1972 (Suppl. II).
26 Jones, M., Conkle, D. M., Ferrans, V. J., Roberts, W. C , Levine, F. H., Melvin, D. B., and Stinson, E. B.: Lesions Observed in Arterial Autogenous Vein Grafts, Circulation 48: 198, 1973 (Suppl. III). 27 Lesperance, J., Bourassa, M. G., Saltiel, J., and Grondin, C. M.: Late Changes in Aorto-Coronary Vein Grafts: Angiographic Features, Am. J. Roentgenol. Radium Ther. Nucl. Med. 116: 74, 1972.