Biomechanics of human common carotid artery and design of novel hybrid textile compliant vascular grafts

B. S. Gupta1,* and V. A. Kasyanov1,2 1 College of Textiles, Department of Textile Engineering, Chemistry, and Science, North Carolina State University, Raleigh, North Carolina 27695-8301; 2Laboratory of Biomechanics, Riga Technical University, Riga, LV-1658, Latvia
The mechanical properties and structure of a human common carotid artery were studied in order to develop criteria for designing and manufacturing compliant textile vascular grafts. The arterial wall comprised a composite of elastin and collagen fibers with the collagen fibers crimped. This structure led to a unique pressurecircumferential stretch ratio curve, the slope of which increased with an increase in strain. The increase in slope was particularly rapid at a stretch ratio above 1.4 or pressure above 120 mmHg. Based on the knowledge gained, a criteria for the design of biomechanically compliant arterial grafts was developed. An elastomeric prestretched polyurethane monofilament yarn with a low modulus of elasticity and a bulked polyester multifilament yarn with a high modulus of elasticity were combined and used as threads in the manufacture of grafts. Tubular structures of diameters in the range 46 mm were made by weaving. Transverse compliance and morphological and permeability properties of these grafts were determined and compared with those of a currently available woven commercial grafts and human carotid arteries. Results indicated that the compliance values of the hybrid grafts were comparable with those of the human carotid artery. Preliminary in vivo studies in dogs showed promising results: a thin, stable neointima developed within 6 months of implantation on the flow surface. 1997 John
Wiley & Sons, Inc.

INTRODUCTION The design and fabrication of synthetic vascular grafts has been a challenging area in vascular surgical research during the past 30 years. Large diameter grafts (diameter 6 mm) used for bypassing arteries in high flow regions such as the thoracic and abdominal aorta have generally performed well. However, the replacement with grafts of small diameter arteries, such as the coronary, renal, and carotid, has not yet been successful and continues to be a problem in reconstructive surgery. Efforts to develop vascular grafts of diameters less than 6 mm with potential for long-term patency have not yet met with success. A major cause for poor performance of such grafts has been shown to be the lack of compliance.14 Replacement of small arteries by rigid woven or knitted prostheses, which have little or no compliance in the circumferential direction, causes the dampening out of the higher harmonic in the pulse wave5 that leads to an increase in the pulse wave velocity and therefore to an increase in wave reflection and energy
*To whom correspondence should be addressed.

loss. The extent to which the pulse amplitude is dampened depends upon the length of the rigid part. For example, in a study by Womersly,5 a 15-cm long rigid section inserted in the femoral artery of a dog showed a reduction of the amplitude by about 13% in the first and 42% in the fourth harmonic. This problem was also noted and discussed by How and Clarke6 (1984). Baird and Abbott7 and Rittgers et al.8 showed that hemodynamic forces play an important role in the formation of thrombus and hyperplastic intima. Doo and colleagues9 determined theoretically and experimentally the differences in the behaviors of an elastic and rigid tube used as a model for an aortic arch. The resulting flow distributions examined showed a difference in the flow behaviors of the rigid and the elastic models of the arterial system. The arterial wall elasticity had an effect on the blood flow distribution; a lack of elasticity led to high turbulence. The work by Stein et al.10 showed that for a given Reynolds number the intensity of turbulence was significantly lower in compliant tubes than in rigid ones. In the latter, unfavorable flow conditions led to the formation of anastomotic aneurysm, development of hyperplastic neointima, and failure of sutures or tearing of the host artery.

Journal of Biomedical Materials Research, Vol. 34, 341349 (1997) 1997 John Wiley & Sons, Inc.

CCC 0021-9304/97/030341-09

342

GUPTA AND KASYANOV

Clearly, successful development of a small diameter vascular graft will depend not only on the use of new biocompatible materials but also on ideas of new constructions. Special attention has been devoted in recent years to the use of polyurethanes, which have good biocompatibility11,12 and deformational and strength properties.6,1316 However, problems have been encountered with the use of small diameter grafts made exclusively of this material. The polymer is found to creep, which leads to the development of aneurysm. Thus, in spite of the availability of convenient spray technology for manufacturing grafts with such elastomers, interest in woven and knitted textile grafts continues because the latter have been successful in medium and large caliber applications, and substantial experience has been gained in their design and construction during the past four decades.1721 The literature makes it clear that mechanical characteristics of vascular grafts play an important role in governing long-term patency. Specifically, the most important consideration to be given in designing a graft is to match the mechanical properties of the prosthesis with those of the host artery. Information exists in the literature on the mechanical properties of the arteries of animals, especially the dog.2226 The work of Hayashi et al. and that of one of the present authors (V.A.K.) and his associates2830 shed some light on the properties of human blood vessels; however, more insight is needed on the structure and properties of arteries before small diameter grafts with maximum potential success can be engineered. The objectives in the present work were to study the structure and mechanical behavior of a human common carotid artery (CCA), and to use the information in designing and constructing a compliant textile vascular graft. The graft so constructed was composite in structure and characterized by nonlinear elasticity and large transverse deformation. Animal studies conducted with the grafts showed highly promising results.

Figure 1. The schematic of the experimental device. 1, sample in chamber with a physiological solution; 2, force transducer; 3, pressure transducer; 4, TV camera; 5, displacement transducer; 6, stepper motor; 7, system for liquid feeding.

MATERIALS AND METHODS Mechanical properties and structure of human CCA Seven human CCAs, retrieved at the autopsies of persons aged 2135 years, were used as experimental materials. The vessels were marked with gentian violet stain before resetting to identify the in situ axial extension ratio. After resection, the specimens were stored in physiological salt solution until the mechanical tests, conducted within 2 h, were performed. The device used for these tests is shown in Figure 1. An artery was cannulated at both ends. The sample was

placed in the chamber with the physiological salt solution maintained at the temperature of 37 1C. One end of the tube was clamped to a support to which a pressure transducer (Micron Inst. M-15) and a specially designed inductive force transducer were connected. The other end was clamped to a support to which a pressure bottle containing fluid was connected. The force transducer recorded the force necessary to maintain the vessel at its in situ length. Axial stretch was introduced by a slide mechanism to which the balance arms were fixed. The axial deformation of the artery was measured with a specially designed inductive strain transducer connected to one of the arms of the balance. Diameter changes in the specimen were sensed optically with a video-dimensional analyzer coupled with a suitable lighting system for high contrast. The changes in diameter with pressure were tracked and recorded continuously. An arterial sample was gradually loaded by internal pressure from 0 to 200 mmHg while maintaining the length of the sample constant at L0, the length in situ. The pressure was elevated in 20-mmHg steps with pressure held constant in each step for 1 min. The initial external diameter at inner pressure p = 0 mmHg and at in situ axial length L0 was noted as D0. The diameter D was recorded at each pressure level. The value of wall thickness h was calculated as follows: h = h0 where
3 3,

(1)

1 1

,
2

(2) (3)

= (D/D0),

and

CAROTID ARTERY/BIOMECHANICS AND NOVEL VASCULAR GRAFTS

343

= (L/L0) = 1.0.

(4)

In these equations, h0 is the initial thickness of the specimen wall and 1, 2, and 3 are, respectively, the stretch ratios in the axial, circumferential, and radial directions. Because the length of the artery was maintained constant at L0, the value of 1 (=L/L0) was 1. The initial wall thickness h0 was measured with a cathetometer to 0.001 mm accuracy. The artery was preconditioned before the tests by subjecting it to cyclic loading to bring it to a stable state, which would give a more reproducible mechanical response. During this process the vessel was pressurized from 0 to 200 mmHg in 20 mmHg steps five times with pressure held constant for 1 min at each step. The initial curves were markedly hysteretic, but the third or fourth cycle gave reproducible curves with minimal hysteresis. The structure of the inner layers was studied after sectioning under a JEM-100C scanning electron microscope (SEM). For this the prepared section was fixed in 3% glutaraldehyde for 8 h, postfixed in 1% osmium tetroxide for 5 h, dehydrated in ethanol of increasing concentration, and dried in a Japanese Eiko-Dh-I critical point apparatus. The dried section was coated in a vacuum chamber with 5060 nm of gold and examined under an SEM with a ASJD-4D scanning attachment at 40-kV accelerating voltage and 80030,000 magnification. Light microscopy (LM) was used to gain preliminary information about the constitutive elements of the connective tissue. For the latter, the samples were fixed in 10% formaldehyde (pH 7.0) and then embedded in paraffin wax, sectioned in various directions, and stained using hematoxylin-eosin. The collagen fibers were dyed using the van Gison method, and the elastin fibers were dyed using the Weigert method.

Manufacturing, mechanical properties, and structure of composite textile grafts The performances of textile composite vascular grafts designed and constructed in this work and of available commercial woven grafts (CVG)1 (6-mm diameter, woven structure made of 7 tex linear density polyester, North, St. Petersburg, Russia) were evaluated. In manufacturing the former, textile threads with two widely different deformative characteristics, one nearly matching those of the elastin and the other of the collagen fibers, were selected. The materials used were a polyurethane monofilament yarn (Volgograd Chemical Thread Plant, Volgograd, Russia) with a low modulus of elasticity (0.8 MPa) and a bulked polyester multifilament yarn (Mogilev Chemical Thread Plant, Mogilev, Russia) with a high modulus of elasticity (1.4 102 MPa). Tubular grafts of diameters 46 mm were

made by a weaving process utilizing a foil ribbon loom. Two types of grafts were made. In the first type (HVG-1), polyester threads of 9 tex linear density were used as the warp (Fig. 2), and the same polyester and a prestretched polyurethane (7.8 tex linear density) were used as the weft. The stretch in the elastic thread caused the crimp to develop in the weft threads, which made the graft stretchable in the transverse direction. In the second variant (HVG-2), prestretched polyurethane thread combined with polyester was used as both the warp and the weft. This combination in the warp threads is novel in that the crimp, usually introduced in the longitudinal direction by the tedious process of crimping and heat setting in commercial grafts, developed automatically in this graft by differential shrinkage. The grafts obtained were stretchable and thus compliant in both the transverse and the longitudinal directions. The mechanical properties of the grafts so produced were determined at pressures ranging from 0 to 200 mmHg by following the procedure described for the carotid arteries. Five specimens of each type were tested. A thin latex tube of diameter larger than those of the grafts was inserted into the specimen before the fluid was passed and the graft pressurized. For determining hydraulic permeability, the procedure of Guidoin et al.31 was used. In this method, the volume of water passing through the wall under a fixed hydrostatic pressure of 120 mmHg was collected for 5 min and expressed as milliliters per minute per square centimeter of water. The grafts were also examined under an SEM for their surface and pore characteristics. Eight HVG-1 hybrid grafts were also implanted in the carotid and femoral regions of mongrel dogs1 for periods of up to 1 year. (In performing the in vivo studies, the guidelines of the Scientific Councils of the Latvian Academy of Sciences and Academy of Medicine were followed.) The form of the pulse wave found in the graft, the healing characteristics of the

Figure 2. The structure of the hybrid textile vascular grafts: (a) the polyester yarn and (b) the polyurethane yarn.

344

GUPTA AND KASYANOV

surface, and the changes in the mechanical properties as a result of implantation were examined.

Compliance and stiffness parameter The flexibility and stiffness of arteries or grafts have been frequently characterized by the values of compliance,32 Cv, pressurestrain elastic modulus,33 Ep, and stiffness parameter,27 . These parameters are defined as follows: Cv = (2/De) ( De/ P); Ep = De ( P/ De); ln (P/Ps) = (D/Ds 1). (5) (6) (7)

In these De is the external diameter at 80 mmHg, P is the pressure difference (12080 mmHg) over which measurements are made, Ps is the mean systemic pressure (100 mmHg), and Ds is the corresponding external diameter. Compliance Cv is thus the fractional change in external diameter, De, with change in physiological pressure, P, from 80 to 120 mmHg (or from 10.67 to 16 kPa), and the coefficient is the slope of the natural logarithm (ln) pressurediameter curve and thus represents the stiffness of the vascular wall. Because the pressurediameter relation of an arterial wall is generally nonlinear, even within the physiological pressure range, the parameters Cv and Ep are not usually material constants but change with the internal pressure. The stiffness parameter , however, is independent of the pressure and has been used to characterize the elastic properties of polyurethane grafts.6,13

Figure 3. Pressure p to circumferential stretch ratio 2 relationship. 1, commercial vascular graft (CVG); 2, hybrid textile vascular graft (HVG-1, after 3 months of implantation); 3, hybrid textile vascular graft (HVG-2); 4, hybrid textile vascular graft (HVG-1); and 5, human common carotid artery.

RESULTS A typical pressure p to circumferential stretch ratio 2 relationship for human CCA, which differs from the relationship generally found on traditional polymer materials, is shown in Figure 3. The p 2 curves for the arteries are concave upward as expected.27,30 The slope gradually increases with stretch ratio and becomes exceptionally high when 2 reaches a value of about 1.42 0.12 (or when pressure exceeds 120 mmHg). The underlying structure of the arterial wall must be responsible for the noted mechanical behavior. Studies by Wolinsky and Glagov34 and Langewouters et al.35 show that the wall tissue is made up of at least two major fibrous materials, an elastin fiber with a low modulus of elasticity (25 105 Pa) and a collagen fiber with a high modulus of elasticity (510 107 Pa). In the relaxed state, the collagen fibers are slack. At low strain or pressure, most of the load is

borne by the elastin fibers, and therefore, the artery is highly distensible. However, as the pressure is raised and the strain is increased, the collagen fibers straighten out and start to bear load. This process causes an increase in the slope. Results from LM and the SEM show that carotid arteries have a specific wavy structure and the tissue is a biocomposite (Figs. 4 and 5). The wavy membranes consist of the elastic fibers plaited with collagen fibers. With an increase of internal pressure the diameter of the artery increases and the degree of waviness decreases [Figs. 4(b) and 5(b)]. At a pressure of 120 mmHg and the longitudinal stretch ratio corresponding to the in situ length, the waviness of the wall elements practically disappears. With a further increase in pressure, the collagen fibers begin to resist the circumferential strain in the artery and thus give rise to the behavior noted in Figure 3. The behavior of grafts HVG-1 and HVG-2, also shown in Figure 3, is similar in character to that of the carotid artery. The polyurethane and polyester threads in the former seem to play about the same roles as the elastin and the collagen fibers, respectively, in the latter. Some differences are noted in the values and the shapes of the curves. A pressure of 160 mmHg or greater is needed in the grafts to straighten out the higher modulus fibers; in arteries the corre-

CAROTID ARTERY/BIOMECHANICS AND NOVEL VASCULAR GRAFTS

345

that HVG-1 grafts are less water permeable than HVG-2 grafts, the values being 2.06 0.16 and 2.42 0.18 mL/min/cm2. These values are about the same as obtained by others on grafts.17,31 HVG-1 grafts of 4-mm internal diameter were implanted in eight mongrel dogs for periods of 1 month to 1 year. Four were implanted in the carotid and the other four in the femoral regions. Six grafts functioned satisfactorily, but the remaining two (femoral) developed a stenose of distal anastomoses. Six months after implantation, the circumferential stretch ratio (for the range 80120 mmHg) decreased from 9.02 to 2.62 (Fig. 3), due obviously to some inward growth of the surrounding tissues. The structure of the grafts was examined with LM and an SEM before implantation and after retrieval following implantation. Figure 6 shows the structure of the outer and inner walls of HVG-1 before implantation. The loops noted are those of the bulked polyester yarns formed by the recovery of the polyurethane yarns. Figures 7 and 8 show typical morphology found in the six patent grafts after implantation. These figures indicate that the healing process pro-

Figure 4. Circumferential histological slice of the carotid arterial wall: (a) at zero internal pressure and (b) at internal pressure of 120 mmHg (sample 9, man 34 years old; original magnification 50).

sponding pressure was lower. For the pressure range 80120 mmHg, the increase in circumferential stretch ratio was 7.14% for the carotid arteries, 9.02% for the HVG-1 graft, 6.06% for the HVG-2 graft, and only 0.32% for the commercial graft. These results show that the commercial grafts used are the least distensible of all materials examined and are not biocompliant. Compliance values of the hybrid textile grafts compare favorably with those of the carotid arteries (Table I). Comparison of the Cv and Ep values shows that the deformability of the HVG-1 graft is greater than that of the carotid artery. Based on the values of the stiffness parameter , one can conclude that graft HVG-1 is more compliant than graft HVG-2, the stiffness coefficients being 4.87 1.56 and 7.81 1.62, respectively. Clearly, the differences in the values of the two grafts arise from the differences in their structures, indicating that a graft with the desired mechanical properties could be constructed if the structure was carefully controlled. The results of the water permeability obtained show

Figure 5. SEM of the circumferential slice of the carotid artery wall: (a) at zero internal pressure and (b) at internal pressure of 120 mmHg (sample 9, man 34 years old; original magnification 5000).

346

GUPTA AND KASYANOV

TABLE I Mean Value SD of Human Common Carotid Artery (CCA), Textile Hybrid Vascular Grafts (HVG-1 and HVG-2), and Commercial Graft (CVG)
Artery or Graft Specimen CCA HVG-1 HVG-2 CVG Wall Thickness (mm) 1.54 0.12 0.42 0.04 1.23 0.06 0.27 0.02 Diameter (mm) 5.92 0.47 4.02 0.04 4.05 0.06 6.01 0.04 Compliance, Cv (kPa1) 0.0238 0.0132 0.0324 0.0083 0.0227 0.0078 0.00186 0.0005 Elastic Modulus, Ep (kPa) 83.86 22.38 61.81 16.92 87.78 19.04 1074.64 136.18 Stiffness Parameter, 5.18 1.94 4.87 1.56 7.81 1.62 102.91 16.84

gressed normally. At retrieval, none of the grafts showed any aneurysm formation, perigraft hematomas, or rupture. All anastomoses were intact, and external tissue reaction was minimal. The inner surface was lined with a smooth, thin layer of transparent glistening tissue, with occasional small foci of yellowtan staining, but there was no evidence of adherent thrombus (Fig. 7). SEM examination showed that the endothelial-like cells appeared flattened with elongated nuclei (Fig. 8). The results of the pulse wave measured on the dogs showed that the pulse wave obtained with compliant grafts was practically of the same form as found with the carotid artery (Fig. 9).

amount of reflected energy can be reduced to zero if a perfect match exists between the fluid impedances of the host artery and the graft. In a study by Scott and Wilson40 in which blood flow behavior in the human leg was simulated with a computer model, it was shown that a match of both the diameters and the compliances of the prosthesis and the host vessel were needed to maximize flow velocity while minimizing reflected energy losses and stresses at the suture lines. The present work shows that in the human CCA, the slope of the pressure (p)circumferential stretch

DISCUSSION Although many factors affect the success of a vascular graft in surgery, the two most important are the mechanical characteristics of the graft and the ability of the graft to heal. The hope for achieving long-term patency in 6-mm grafts lies in matching the compliance of a vascular graft with that of the artery and developing a thromboresistant surface at the luminal wall. Numerous studies have shown that a positive correlation exists between the matching of compliance and the patency of the graft.1,2,12,14,36,37 Preliminary studies show15 that compliance is a particularly important factor during the first six or so weeks of the operation; long-term compliance may be a negligible factor in determining the overall patency of small grafts. In an ideal graft, the velocity of flow is high while the stresses at the suture lines and the reflected energy losses are small. For achieving minimum stress concentration and energy losses in the existing commercial grafts, the diameter of the graft chosen has to be 1.41.5 times the diameter of the host artery.38 An impedance mismatch between the host and the graft leads to development of eddy currents and wave reflections. These contribute to structural fatigue, false aneurysm, thrombus formation, hyperplastic tissue growth, and atherosclerotic changes in the host artery, which can lead to early occlusion of the graft.14,39 The

Figure 6. SEM of the hybrid textile vascular grafts: (a) outside structure and (b) inside structure.

CAROTID ARTERY/BIOMECHANICS AND NOVEL VASCULAR GRAFTS

347

Figure 7. Hybrid vascular graft HVG-1 after 6 months of implantation. The interior is lined with a thin layer of transparent glistening tissue without thrombus.

ratio ( 2) curve increases with pressure and becomes exponential at pressures greater than 120 mmHg (Fig. 3). The LM and SEM analyses show that the arterial wall is a biocomposite of elastin and collagen fibers and has a specific crimped character. This structure, as explained earlier, leads to the unusual mechanical behavior noted above, which differs from that of the traditional polymer materials. The value of compliance measured on human carotid arteries coincides with the values obtained by Hayashi et al.27 who found that the stiffness parameter had a value of 5.25 for the human CCA and 19.84 for the femoral artery. Our results show a value of 5.18 for the human carotid artery (see Table I). The novel hybrid textile vascular grafts produced in this work have elastic properties that match those of the human carotid artery. The materials used in their construction were polyurethane and polyester yarns that allowed the hybrid textile grafts to have a compliance value in the circumferential direction 10 or more times those of the currently available commercial grafts. The water permeability value of the hybrid grafts is at the same level as found in the commercial prostheses. All commercial grafts are crimped by one or two heat-setting processes. These grafts rapidly lose crimp and longitudinal compliance upon implantation.17 The procedure used in manufacturing the hybrid textile prostheses, on the other hand, renders the heat setting process for crimping unnecessary. The combination of prestretched polyurethane yarns with polyester in the warp threads led to the development of a stable crimp in the longitudinal direction. Moreover, by varying the sizes and properties of the individual yarns, the structure and properties of vascular grafts could be effectively and conveniently engineered to suit the application. However, the mechanical compliance alone does not guarantee success in vivo. Kambic et al.41 showed that no one material satisfies the requirements for by-

pass of all small caliber arteries. Future research must additionally focus on development of improved inner surface coatings, which may minimize tissue reaction and undesirable cellular events at the anastomoses. A gelatin coating was used to provide a bloodcompatible graft with a smooth nonpseudoneointima generating surface that does not promote cell ingrowth at the anastomoses.15 A difficult problem faced in matching compliance is a change in the mechanical properties of the graft that is induced by tissue ingrowth. The in vivo test results on the grafts of this study are encouraging. Examination of the HVG-1 grafts after 6 months of implantation indicated that the healing process progressed normally. Some loss of compliance was noted due to tissue ingrowth into the wall. Hasegava and Azuma,42 working with woven Dacron or Teflon grafts, found that 3 weeks postimplantation, the longitudinal stiffness had increased but the circumferential stiffness had not changed. On the other hand, the work of Lee and Wilson37 showed a marked increase in circumferential stiffness after 3 months of implantation, indicating that at 3 months the connective tissue ingrowth was organized and played a significant role in increasing wall stiffness. It was also found43 that although both neointima and adventitia cells were closely attached to the polyurethane fibers near the surface, there was no transmural or through growth of the tissues. After 9 months of implantation in mini-pigs, the grafts were still fairly compliant. Studies on spandex prostheses in dogs indicated18 that while these grafts lost some circumferential stretchability after becoming infiltrated with unyielding collagenous tissue, they remained compliant during the first week after insertion and adapted their diameter to the flow conditions of the arteries that were bypassed. The spandex grafts usually adjust their diameter in response to changing flow

Figure 8. Endothelium cells on the flowing surface of the HVG-1 graft.

348

GUPTA AND KASYANOV

References
1. I. G. Kidson and W. M. Abbot, Low compliance and arterial graft occlusion, Trans. Am. Soc. Artif. Intern. Organs, 58, 1.11.4 (1978). 2. R. Walden, G. J. ILtalien, J. Megerman, and W. M. Abbott, Matching elastic properties and successful arterial grafting, Arch. Surg., 115, 11661169 (1980). 3. J. G. Sladen and T. M. Maxwall, Experience in vivo 130 polytetrafluoethylene grafts, Am. J. Surg., 141, 545548 (1981). 4. R. A. White, The effect of porosity and biomaterial on the healing and long-term mechanical properties of vascular prostheses, Trans. Am. Soc. Artif. Intern. Organs, 34, 94100 (1988). 5. J. R. Womersly, Oscillatory flow in arteries. II: The reflection of the pulse wave at junction and rigid inserts in the arterial system, Phys. Med. Biol., 2, 313323 (1958). 6. T. V. How and R. M. Clarke, The elastic properties of a polyurethane arterial prosthesis, J. Biomech., 17, 597 608 (1984). 7. R. N. Baird and W. M. Abbott, Pulsatile blood flow in arterial grafts, Lancet, 2, 948950 (1976). 8. S. E. Rittgers, P. E. Karayannacos, J. F. Guy, R. M. Nerem, G. H. Shaw, J. R. Hostetler, and J. S. Vasko, Velocity distribution and initial proliferation in autogenous vein grafts in dogs, Circ. Res., 42, 792801 (1978). 9. I. Doo, W. Jedruch, J. Kennedy, P. Adams, and C. M. Rodkiewicz, Wall distensibility effect on arterial flow distribution, J. Biomech., 17, 643 (1984). 10. P. D. Stein, F. J. Walburn, and E. F. Blick, Damping effect of distensible tubes on turbulent flow: implication in cardiovascular system, Biorheology, 17, 275281 (1980). 11. P. Beahan and D. Hull, A study of the interface between a fibrous polyurethane arterial prosthesis and natural tissue, J. Biomed. Mater. Res., 16, 827838 (1982). 12. K. Hayashi, Keynote lecture. Biomechanical approach to the design of vascular prostheses, Tissue Eng., 14, 510 (1989). 13. K. Hayashi, K. Takamizawa, T. Saito, K. Kira, K. Hiramatsu, and K. Kondo, Elastic properties and strength of a novel small-diameter, compliant polyurethane vascular graft, J. Biomed. Mater. Res. Appl. Biomater., 23, 229244 (1989). 14. B. van der Lei and Ch. R. H. Wildevuur, From a synthetic, microporous, compliant, biodegradable smallcaliber vascular graft to a new artery, Thorac. Cardiovasc. Surg., 37, 337347 (1989). 15. H. E. Kambic, Polyurethane small artery substitutes, Trans. Am. Soc. Artif. Intern. Organs, 34, 10471050 (1989). 16. G. Soldani, G. Panol, H. F. Sasken, M. B. Goddard, and P. M. Galletti, Small diameter polyurethanepolydimethylsiloxane vascular prostheses made by a spraying, phase-inversion process, J. Mater. Sci. Mater. Med., 3, 106, 113 (1992). 17. M. W. King, R. G. Guidoin, K. R. Gunasekera, and C. Gosselin, Designing polyester vascular prostheses for the future, Med. Progr. Technol., 9, 217226 (1983). 18. N. Rosenberg, A. N. Simpson, and R. E. Brown, A circumferential elastic prostheses: three-year studies of a dacron-spandex graft in the dog, J. Surg. Res., 34, 716 (1983). 19. C. C. Chu and L. E. Lecaroz, Design and in vivo testing

Figure 9. Shape of the pulse waves: (a) dog carotid artery, (b) commercial graft, and (c) hybrid vascular graft HVG-1.

and pressure conditions, particularly the former, and develop a thin, almost transparent, neointima on the flow surface. Another important parameter that determines the success of a graft in bypass application is porosity. Tissue incorporation and healing of synthetic grafts are related to this property.44 The pore size affects the type of tissue grown through the wall.4 Working with polyurethane grafts, White4 showed that if the pore size was less than 15 m, minimal tissue ingrowth took place; if it was greater than 15 m but less than about 45 m, fibrohistiocytic tissues grew; and if the size was greater than about 50 m, the structure was infiltrated with organized fibrous tissue. The above results thus indicate that by choosing appropriate materials and controlling the structure, a textile graft can be engineered that has the desired compliance and the potential for attaining the needed tissue ingrowth. An important question still remains; however: Should the properties of the host be matched by those of the virgin prostheses or those of the prostheses after they have been at the site for some length of time?45 To address this question, it is necessary to know what changes take place in the properties of a graft with time in situ and how these affect the function of the product. Clearly, this knowledge is needed before an optimally useful graft can be designed.
The benefits accruing to the authors from a commercial or industrial party will be applied to a research fund, nonprofit institution, or other organization with which the authors are associated.

CAROTID ARTERY/BIOMECHANICS AND NOVEL VASCULAR GRAFTS

349

20. 21.

22. 23. 24. 25. 26. 27.

28.

29. 30. 31.

32. 33.

of newly made biocomponent knitted fabric for vascular surgery, Adv. Biomed. Polym. Proc. Symp., Chicago, Sept. 815, 1987. L. ShuTung, A collagen-dacron composite vascular graft for arterial reconstructions, Adv. Biomed. Polym. Proc. Symp., Chicago, Sept. 815, 1987. V. Kasyanov and B. Purinya, Elastic properties of the composite vascular grafts, presented at the 3rd International ITV Conference on Biomechanics, Medical Textiles for Implantation, Stuttgart, Germany, 1989, p. 39. R. H. Cox, Anisotropic properties of the canine carotid artery in vitro, J. Biomech., 8, 293300 (1975). D. J. Patel and D. L. Fry, The elastic symmetry of arterial segments in dogs, Circ. Res., 24, 18 (1969). D. J. Patel and J. S. Janicki, Statical elastic properties of the left coronary circumflex artery and common carotid artery in dogs, Circ. Res., 27, 149158 (1970). R. N. Vaishnav, J. T. Young, J. S. Janicki, and D. J. Patel, Nonlinear anisotropic elastic properties of the canine aorta, Biophys. J., 12, 10081027 (1972). P. B. Dobrin, Biaxial anisotropy of carotid artery: estimation of circumferential elastic modulus, J. Biomech., 19, 351358 (1986). K. Hayashi, H. Handa, S. Nagasawa, A. Okumura, and K. Moritake, Stiffness and elastic behavior of human intracranial and extracranial arteries, J. Biomech., 13, 175184 (1980). V. Kasyanov and A. Kregers, Differences in the deformation and strength properties of human large blood vessels as a function of location, loading direction and age, Polym. Mech., 4, 606611 (1975). B. Purinya and V. Kasyanov, Changes in the mechanical properties of human coronary arteries with age, Polym. Mech., 2, 251255 (1977). V. Kasyanov, B. Purinya, and E. Ceders, Determination of the shear modulus of human blood vessel wall, Polym. Mech., 5, 753755 (1978). R. Guidon, M. King, D. Marceau, A. Cardou, D. Faye, J.-M. Legendre, and P. Blais, Textile arterial prostheses: Is water permeability equivalent to porosity, J. Biomed. Mater. Res., 21, 6567 (1987). B. S. Gow and M. G. Taylor, Measurement of viscoelastic properties of arteries in the living dog, Circ. Res., 23, 111122 (1968). L. H. Peterson, R. E. Jensen, and R. Parnell, Mechanical properties of arteries in vivo, Circ. Res., 8, 622639 (1960).

34. H. Wolinsky and S. Glagov, Structural basis for the static mechanical properties of the aortic media, Circ. Res., 14, 400413 (1964). 35. G. J. Langewouters, K. H. Wesseling, and W. J. Goedhard, The pressure dependent dynamic elasticity of 35 thoracic and 16 abdominal human aortas in vivo described by a five component model, J. Biomech., 18, 613620 (1985). 36. D. J. Lyman, D. Albo, Jr., R. Jackson, and K. Knutson, Development of small diameter vascular prostheses, Trans. Am. Soc. Artif. Intern. Organs, 23, 253261 (1977). 37. J. M. Lee and G. J. Wilson, Anisotropic tensile viscoelastic properties of vascular graft materials tested at low strain rates, Biomaterials, 7, 423431 (1986). 38. P. E. Paasche, C. E. Kinley, F. G. Dolan, E. R. Gonza, and A. E. Marble, Consideration of suture line stresses in the selection of synthetic grafts for implantation, J. Biomech., 6, 253259 (1973). 39. R. R. Kowligi, W. W. von Maltzahn, and R. Eberhart, Fabrication and characterization of small-diameter vascular-prostheses, J. Biomed. Mater. Res., 22, 245256 (1988). 40. S. M. Scott and B. S. Wilson, The mechanical design of vascular prostheses, presented at The Symposium on Hemodynamics of the Limbs, Toulouse, France, May 17 19, 1979, pp. 251259. 41. H. Kambic, S. Murabayashi, and Y. Nose, Biomaterials in artificial organs, Chem. Eng. News, 64, 3048 (1986). 42. M. Hasegava and T. Azuma, Mechanical properties of synthetic arterial grafts, Biomechanics, 12, 509517 (1979). 43. D. Annis, A. Bornat, R. Edwards, A. Higham, B. Loveday, and J. Wilson, An elastomeric vascular prosthesis, Trans. Am. Soc. Artif. Intern. Organs, 24, 209214 (1978). 44. S. A. Wesolowski, C. C. Fries, R. T. Domingo, W. I. Liebig, and P. N. Sawyer, The compound prosthetic vascular graft: a pathologic survey, Surgery, 53, 1944 (1963). 45. T. V. How, Mechanical properties of arteries and arterial grafts, in Cardiovascular Biomaterials, G. W. Hastings (ed.), SpringerVerlag, New York, 1992, pp. 135. Received February 6, 1995 Accepted March 8, 1996