The generation of a small-caliber vascular graft prosthesis is a challenge because of its poor long-term patency.1,2 For example, some of the synthetic materials pose a risk of infection and inflammatory response and/or lack the capacity for growth or auto tissue replacement after bypass grafting. Recently, tissue engineering of cardiovascular prostheses has been receiving attention, because it is expected to overcome such vascular prosthesis problems.
Despite high expectations for the development of artificial scaffold materials for vascular tissue engineering, optimal synthetic materials have not yet appeared.3 On the other hand, decellularized matrices have been used for vascular tissue engineering with relatively good prognoses.4–8 One of the reasons for this enhanced prognosis is the superior biocompatibility of the decellularized matrix compared with that of the synthetic scaffold materials. In particular, decellularized matrices have proved better at facilitating cell attachment and recellularization.5,8 Another reason might be their mechanical properties. Although vascular smooth muscle cells play an important role in vascular contraction and relaxation, the importance of the extracellular matrix (ECM) in the maintenance of the mechanical properties of the vessels has been suggested.6 Some studies have reported that decellularized vessels used as scaffolds for tissue engineering of small-caliber vascular grafts showed similar results in the isometric tension examination5 and in the uniaxial stress-strain response8 after bypass grafting compared with those of native arteries. However, the detailed compliance of the decellularized matrices is not well documented.
Compliance is an important factor in designing the scaffold for tissue-engineered small-caliber vascular grafts, since a compliance mismatch is thought to affect the prognosis. Such a mismatch causes neointimal hyperplasia at anastomosis sites, which, in turn, leads to thrombosis and occlusion even at an early stage after bypass surgery.9–12 The mechanical properties of the graft are largely affected by the materials used as a scaffold, at least in the initial phase after grafting, making the evaluation of the scaffold compliance very important. Preoperative compliance of the decellularized vessels has not been sufficiently evaluated, especially compliance such as that involved in cross-sectional area changes in response to the intraluminal pressures. Furthermore, most of the studies that have evaluated compliance of the native vessels or tissue-engineered vascular grafts, measured external diameter changes.6,13,14 However, the measurement of internal diameters is important to eliminate the influence of the wall thickness of the scaffold.15 Furthermore, cross-sectional areas are probably more appropriate than internal diameters in a compliance evaluation of tubular conduits such as vessels, since the vessels are not always completely circular.
Intravascular ultrasound (IVUS) is now widely used in angioplasty to measure internal diameters and cross-sectional areas by observing the cross-sectional aspects of the vessels.16 It has been also used to evaluate the mechanical properties of the vessels in human coronary17 and brachial arteries18in vivo. If IVUS is adopted as the evaluation system of a vascular graft, compliance can be easily assessed by measuring the internal diameters and the cross-sectional areas. In particular, IVUS does not require contrast media, which is advantageous in the evaluation of a tissue-engineered vascular graft. Furthermore, IVUS allows observation of the vascular walls aseptically during culture in a bioreactor. Because preevaluation of the tissue-engineered vascular graft is extremely important to ensure the safety of the operation, this system should prove beneficial in future clinical trials.
In this study, we measured the compliance and stiffness of candidate scaffolds for vascular tissue engineering, decellularized vessel and ureter, and elastin-added synthetic material by using IVUS, and compared them with those of native compliant artery and noncompliant synthetic material to better understand the detailed characteristics of decellularized tissue.
Materials and Methods
Selection of Materials to Evaluate Compliance
Five materials were selected for evaluation: Polytetrafluoroethylene (PTFE) (Gore-Tex, W. L. Gore & Associates, Inc., Newark, DE) was selected as a conventional noncompliant vascular graft. Canine common carotid artery (CCA) was selected as an example of the compliant native artery. CCA and canine ureter decellularized by deoxycolic acid (DCCA and D. Ureter) were evaluated to investigate the mechanical properties of decellularized scaffolds. It was also of interest to us to evaluate the contribution of cell components to vascular compliance. As a suitable material to investigate the influence of elastin-added synthetic material, we proposed an elastin gel combined with a polylactic acid nanofiber nonwoven tube (E-PLA) created by electrospinning as a biodegradable elastic vascular scaffold19 (kindly provided by the Department of Tissue Engineering Development, Innovation Research Institute, Teijin Limited, Tokyo, Japan).
Preparation and Decellularization of Canine CCA and Ureter
CCA and ureter were harvested from beagle dogs (mean body weight, 11 kg) when they were euthanized for other experiments. The CCA and ureter were decellularized by deoxycolic acid according to the method described previously with minor modifications.20 Briefly, two CCAs or ureters were treated with 30 ml of phosphate-buffered saline (PBS) solution containing 0.1% sodium azide (Wako Pure Chemical Industries, Osaka, Japan) in a 50-ml PP test tube (Greiner Bio-One, Frickenhausen, Germany) and agitated on a shaker (Labo Shaker, BC-730, Biocraft, Tokyo, Japan) for 12 hours at the maximum frequency of 140/min. Next, the samples were treated with 30 ml of PBS solution supplemented with 5.8% NaCl, 0.02% DNase (Roche, Basel, Switzerland), and 0.01% RNase (Roche) in a new PP test tube for 12 hours. They were then treated twice with 30 ml of PBS solution containing 4% deoxycolic acid (Sigma, St. Louis, MO) in a new PP test tube for 12 hours. After this procedure, the samples were put in another new PP test tube with 30 ml of PBS with 1% antibiotic-antimycotic reagent (Invitrogen Corporation, Carlsbad, CA) and washed under a shaker for 3 days, changing the PBS solution every 12 hours. When this process was completed, excess remnant tissue was removed with a forceps. Decellularization was confirmed by hematoxylin and eosin staining. The remaining DNA content of each decellularized method was compared in the other experiment by using esophagi.21 The decellularized tissues were lyophilized, and DNA isolation was performed with a Tissue DNA Isolation kit (PUREGENE, Minneapolis, MN), according to the manufacturer's protocol. The results showed superior decellularization with the use of deoxycholic acid.
All animal experiments were performed in accordance with the Guidelines for Animal Experimentation of the Nagoya University School of Medicine.
Generation of Evaluation System for Small-Caliber Vascular Compliance With the Use of IVUS
As shown in Figure 1, the system was equipped with metal connectors for the vascular prosthesis in a closed circuit and connected via side branches to a pressure monitor (78342A Monitor, HP, CA) and a syringe pump (TE-331, Terumo, Tokyo, Japan). A vascular graft prosthesis, approximately 10 cm in length, was tied to the metal connectors with surgical threads. The circuit was filled with Klebs-Ringer solution (Sigma). A 6F coronary guiding catheter (Judkins Right, Medtronic, Minneapolis, MN) was inserted through an 8F sheath attached to the end of the circuit. Another guiding catheter was also inserted from the other end of the circuit. A guide wire (BMW, Guidant, Indianapolis, IN) was inserted from one guiding catheter that also penetrated the end of the other catheter. The guide wire was secured tightly to keep it in a coaxial cross-sectional position during the experiment. An IVUS probe (Atlantis Plus, Boston Scientific, Boston, MA) was inserted along with the guide wire into the material, enabling an accurate view of a cross-sectional image. To avoid the influence of end effects, the IVUS probe was placed in the middle of the vascular graft. The distance between the ends of the material was 8 cm, and the IVUS probe was placed at the points at least 3 cm from the end of the connecting surface. By changing the infusion flow rate of the syringe pump, we could generate a constant intraluminal pressure on the material every 5 mm Hg, from 0 to 200 mm Hg.
Measurement of Intraluminal Pressure, Internal Diameter, and Cross-Sectional Area
Internal diameters of the vessels were measured for two directions (vertical and horizontal in relation to the ground) at each pressure point (41 points were measured between 0 and 200 mm Hg). The longest line in each direction was considered as diameter. The measurements were performed by manual tracing, using cross-sectional images generated by Scion Image Software (Scion Corporation, Frederick, MD). Three consecutive experiments were performed. In each experiment, three different locations for each material were measured at each pressure, and the results were expressed as mean value ± standard deviation.
Evaluation of Material Compliance
We evaluated the pressure–cross-sectional area relation (P-A relation),17 pressure percent area change relation (P-%AC relation), stiffness parameter β (β value),13 and diameter compliance (Cd)11 to compare the compliance and stiffness among candidate scaffolds. Definitions for each parameter used in this study were as follows. The P-%AC relation is calculated to normalize the cross-sectional area changes from multiple data points in the material. This parameter is considered to represent a more accurate relation than the pressure-% diameter change relation described previously,6 since the cross-sectional area was used instead of a simple diameter observable only in one direction. The percent area change relative to the area at the standard intraluminal pressure (Ps) is calculated as
where ΔS is the percent area change, Sp is the cross-sectional area at a target pressure, and Ss is the reference cross-sectional area at Ps. The β value, which was proposed by Hayashi et al.,13 is determined according to the following equation:
where P is the pressure and D is the internal diameter at P, and Ps and Ds are the standard intraluminal pressure and the internal diameter at Ps, respectively. This parameter is defined by the linear relations observed between the logarithm of pressure ratio, ln(P/Ps), and distention ratio of diameter, D/Ds, in the physiological pressure range from 60 to 160 mm Hg with high correlation coefficient.13 A smaller β value indicates a more compliant material texture. The diameter compliance (Cd) is calculated using the formula
where D is the internal diameter of the vessel, and ΔD is the diameter change associated with ΔP (pressure increment) at the intraluminal pressure of Ps. The unit is expressed as the percentage diameter change per mm Hg × 10–2. In the calculation of percent area change and Cd, we adopted the criteria by which the standard pressure (Ps) was 100 mm Hg and the pressure change (ΔP) was 10 mm Hg, as they were both conventionally performed.6,11
To evaluate the difference between the cross-sectional area measurement and the internal diameter measurement in compliance evaluation, the difference in internal diameters at right angle were compared. The calculated differences between horizontal (a) and vertical (b) diameter lengths at each pressure point were measured with CCA as an example of compliant material and also with PTFE as an example of the noncompliant material. The average b/a ratios of three consecutive measurements were calculated for each pressure point. If the cross-sectional aspect is a complete circle, b/a = 1.
For the analysis of differences in horizontal and vertical diameters, the paired t test was used. Values of p < 0.05 were considered statistically significant.
With IVUS images of CCA, the borderline of the inner surface of the vessel was clearly observed (Figure 2). The internal diameter and cross-sectional area were increased according to the increase in the intraluminal pressure. When the P-A relation of each material was estimated, PTFE showed a noncompliant response and almost no cross-sectional area change (Figure 3). CCA and DCCA showed compliant responses and the graph presented a J-shaped curve, which represent a marked increase in distensibility at the lower pressure range, followed by a gradual reduction in pressure-dependent distensibility at the physiological pressure range, with little inflation observed in the high pressure range.15 E-PLA also showed an increase in the cross-sectional areas according to the pressure load, although the increases were much smaller than CCA, DCCA, or D. Ureter. D. Ureter also showed a compliant response like the J-shape curve, in which cross-sectional area change was larger than that in CCA or DCCA in most of the pressure ranges.
P-%AC relations were evaluated to compare the compliance of each material by normalizing the cross-sectional area changes as the percentages from the areas at 100 mm Hg (Figure 4). These normalized relations represent a clearer J-shaped curve response than described in the P-A relation. CCA and DCCA showed a similar percent area change for every pressure. The cross-sectional area changes of E-PLA also decreased according to the high pressure load, which was clearly observed by normalization of the data from multiple data points in the high pressure range. However, the response was different from the J-shaped curve observed in CCA and DCCA. Actually, the P-A relation curve of E-PLA could be approximated by the linear, positive proportional response (y = 74.3x – 620.3, r = 0.98). D. Ureter showed a similar percent area change in relation to E-PLA from the physiological to the high pressure range (from 60 to 200 mm Hg). However, its compliant response was similar to that of DCCA in the low pressure range (<60 mm Hg) and was a markedly greater percent area change than any other material <30 mm Hg.
We also evaluated the β values and Cds of each material (Table 1). The mean β and Cd values for PTFE (141.3 and 0.83%/mm Hg × 10–2, respectively) and those for CCA (7.8 and 12.2%/mm Hg × 10–2, respectively) were comparable to those in previous reports (β and Cd values for PTFE were ∼150 and ∼1.6%/mm Hg × 10–2, and those for CCA were 5.25 to 40, and 5.5% to 9.8%/mm Hg × 10–2, respectively).10,11,13,15 D. Ureter showed the smallest β value of all materials. With this parameter, D. Ureter was the most compliant, but Cd was also smaller than CCA and DCCA.
The difference between the horizontal and vertical diameter lengths was calculated at 41 pressure points. The vertical diameters were significantly longer than horizontal diameters in both CCA and PTFE (p < 0.0001) (Figure 5, A and B). The mean relative diameter ratio (b/a) for CCA and PTFE was 1.05 ± 0.02 and 1.03 ± 0.008, respectively.
Creating a compliant scaffold similar to native arteries is a challenge because of their specific and complex ECM structure. One way to use a compliant scaffold similar to arteries is decellularization of the native arteries. In this study, we used IVUS to evaluate the vascular material compliance in a static pressure load system.
We selected DCCA to evaluate the contribution of ECM and cell components such as vascular smooth muscle cells and fibroblasts to vascular compliance. DCCA showed a response similar to CCA in its compliance and a stiffness parameter. Although its internal diameter was increased by decellularization, DCCA showed almost identical compliance to that in CCA, as was clearly confirmed by the normalization of cross-sectional areas at 100 mm Hg. These results suggest that cell components do not play a major role in vascular compliance under a static mechanical pressure load and that the complex tubular structure of ECM such as CCA is important to produce a J-shaped curve compliance.
We also selected E-PLA to evaluate the compliance of the material with artificially added elastin. E-PLA showed cross-sectional area changes according to the intraluminal pressure increments, which were quite different from PTFE. The total cross-sectional area change was much smaller compared with that in CCA or DCCA, which probably shows the limitations of currently available biodegradable scaffolds. The difference in cross-sectional area change between the synthetic and the decellularized matrix may also reflect a difference in structure, since the decellularized natural scaffolds are composed of multiple layers. For example, the walls of CCA and DCCA consist of intima, internal elastic lamina, media, external elastic lamina, and adventitia, each of which plays a distinctive role. This complex structure may contribute to developing physiological compliance as in the J-shaped curve. On the other hand, E-PLA has a homogeneous structure that affects the compliance of the scaffold. The difference was evident especially at the lower pressure range compared with CCA and DCCA, where the P-A relation curve of E-PLA could be approximated as a linear response.
D. Ureter had a compliant response that was not completely the same as that in DCCA, in that it showed a much greater increase in the change of the cross-sectional area than DCCA, and that increase was reduced over 80 mm Hg. This may be due to the difference in the ECM structure and environment in vivo between artery and ureter. The wall of the ureter also has multiple layers consisting of transitional epithelium, a mucosal layer, two different directional smooth muscle cell layers, and an outer membrane. The layered structure observed in DCCA and D. Ureter probably contributes to the compliant response shown in the P-A relation. Although these two organs are tubular conduits that transport blood or urine, their environments in vivo are quite different. Blood flows regularly through the artery by the action of the cardiac pump, whereas urine is transported through the ureter by peristalsis. The artery is exposed to mechanical stress and periodically to systolic and diastolic pressure. On the other hand, the ureter is distended intermittently when the urine is transported. These physiological differences may explain the difference in the P-A relation between DCCA and D. Ureter. Furthermore, the subepithelial mucosal layer of the ureter is thicker than the subintimal media of the artery. This difference in the subepithelial middle layer thickness between the ureter and artery may contribute to the difference in the increase of cross-sectional areas in the lower pressure range (Figure 3). However, the results of the present study suggest that the mechanical properties of D. Ureter were much more like those of the DCCA than those of the artificial materials used in this study. Interestingly, D. Ureter showed cross-sectional area changes similar to those of DCCA according to intraluminal pressures, and it could tolerate over 200 mm Hg despite the low pressure environment of the ureter in vivo. These facts support the notion that D. Ureter may have potential as a compliant allovascular or zenovascular graft, although there are some differences in mechanical properties compared with those of DCCA.
There have been several reports on the use of a vascular xenograft made of decellularized bovine ureter (SynerGraft) in animal experiments.22,23 This graft showed a good patency in canine abdominal aorta replacement during 10 months of follow-up. One study reports rapid recellularization by recipient cells and no adverse effects, including aneurysmal formation.22 However, a cautionary note was sounded in a report of a human femoral-posterior tibial bypass operation.24 The patient presented at 8 weeks with erythematous swelling that increased in size along the course of the graft, and the explanted graft showed areas of aneurysmal formation. Considering these reports, further studies are needed to confirm the utility and safety of decellularized zenografts for human clinical applications.
IVUS images revealed that the vascular cross-sectional aspects were not complete circles, as often observed in percutaneous coronary intervention. IVUS is performed along with a guide wire, which is inserted into the guiding catheter. Accordingly, the end of the guide wire usually has no support.16–18 In this study, we used two guiding catheters at both ends of the circuit to remove the instability of the guide wire, so that constant coaxial cross-sectional images can be obtained, even though vertical internal diameters were significantly longer than horizontal internal diameters. This might be partly due to the influence of gravity for the intraluminal Klebs-Ringer solution, and this could be true for native arteries in vivo. This result suggests that the measurement of the cross-sectional area changes might be more beneficial to avoid possible fluctuation of the data caused by horizontal and vertical deviation, especially when the measurement is required before and after bypass grafting.
Although the evaluation system using IVUS in the present study can allow us to evaluate the compliance of the tissue-engineered vascular graft in a bioreactor, we should keep in mind that the flow characteristics can be influenced by the probe in place. It would be appropriate to evaluate at a limited time point (possibly before or after cell culture) when this system is applied for the bioreactor.
In conclusion, a decellularized canine artery, at this time, is the scaffold most similar to the native canine artery in its compliance and stiffness. The main advantage in producing desirable mechanical properties lies in the ECM structure. Decellularized canine ureter may have potential as an allo- or zeno-compliant scaffold, but further studies and techniques are needed to confirm its usefulness and safety. In tissue engineering of a small-caliber vascular graft, the evaluation of the material's mechanical properties, including its compliance before bypass grafting, is necessary to guarantee the safety of the operation
This study was supported in part by a Grant-in-Aid for Industry University and Officer Co-operation Innovation from the Ministry of Education, Culture, Sports, Science, and Technology and a grant from the Tissue Engineering Initiative Incorporation. The authors are grateful to Eiichi Kitazono, Hiroaki Kaneko, and Yoshihiko Sumi (Tissue Engineering Laboratory, Innovation Research Institute, Teijin Limited, Tokyo, Japan) for providing the E-PLA.
1.Teebken OE, Haverich A: Tissue engineering of small diameter vascular grafts. Eur J Vasc Endovasc Surg
23: 475–485, 2002.
2.Edelman ER: Vascular tissue engineering designer arteries. Circ Res
85: 1115–1117, 1999.
3.Xue L, Greisler HP: Biomaterials in the development and future of vascular grafts. J Vasc Surg
37: 472–480, 2003.
4.Wilson GJ, Courtman DW, Klement P, et al: Acellular matrix: a biomaterials approach for coronary artery and heart valve replacement. Ann Thorac Surg
60: S353–S358, 1995.
5.Kaushal S, Amiel GE, Guleserian KJ, et al: Functional small-diameter neovessels created using endothelial progenitor cells expanded ex vivo. Nat Med
7: 1035–1040, 2001.
6.Conklin BS, Richter ER, Kreutziger KL, et al: Development and evaluation of a novel decellularized vascular xenograft. Med Eng Phys
24: 173–183, 2002.
7.Schaner PJ, Martin ND, Tulenko TN, et al: Decellularized vein as a potential scaffold for vascular tissue engineering. J Vasc Surg
40: 146–153, 2004.
8.Borschel GH, Haung YC, Calve S, et al: Tissue engineering of recellularized small-diameter vascular grafts. Tissue Eng
11: 778–786, 2005.
9.Abbott WM, Megerman J, Hasson JE, et al: Effect of compliance mismatch on vascular graft patency. J Vasc Surg
5: 376–382, 1987.
10.Walden R, L'Italien GJ, Megerman J, Abbott WM: Matched elastic properties and successful arterial grafting. Arch Surg
115: 1166–1169, 1980.
11.Hasson JE, Megerman J, Abbott WM: Increased compliance near vascular anastomoses. J Vasc Surg
2: 419–423, 1985.
12.Weston MW, Rhee K, Tarbell JM: Compliance and diameter mismatch affect the wall shear rate distribution near an end-to-end anastomosis. J Biomech
29: 187–198, 1996.
13.Hayashi K, Hanada H, Nagasawa S, et al: Stiffness and elastic behavior of human intracranial and extracranial arteries. J Biomech
13: 175–184, 1980.
14.Niklason LE, Abbott W, Gao J, et al: Morphologic and mechanical characteristics of engineered bovine arteries. J Vasc Surg
33: 628–638, 2001.
15.Sonoda H, Urayama S, Takamizawa K, et al: Compliant design of artificial graft: compliance determination by new digital x-ray imaging system-based method. J Biomed Mater Res
60: 191–195, 2002.
16.Nissen SE, Yock P: Intravascular ultrasound: novel pathophysiological insights and current clinical applications. Circulation
103: 604–616, 2001.
17.Williams MJA, Stewart RAH, Low CJS, Wilkins GT: Assessment of the mechanical properties of coronary arteries using intravascular ultrasound: an in vivo study. Int J Card Imaging
15: 287–294, 1999.
18.Bank AJ, Wilson RF, Kubo SH, et al: Direct effects of smooth muscle relaxation and contraction on in vivo human brachial artery elastic properties. Circ Res
77: 1008–1016, 1995.
19.Kitazono E, Kaneko H, Miyoshi T, Miyamoto K: Tissue engineering using nanofiber. J Synth Org Chem Jpn
62: 514–519, 2004.
20.Selahittin C, Christopher C, Norbert S, et al: The bladder acellular matrix graft in a rat chemical cystitis model: functional and histologic evaluation. J Urol
168: 798–804, 2002.
21.Ozeki M, Narita Y, Kagami H, et al: Evaluation of decellularized esophagus as a scaffold for cultured esophageal epithelial cells. J Biomed Mater Res Part A
22.Charke DR, Lust RM, Sun YS, et al: Transformation of nonvascular acellular tissue matrices into durable vascular conduits. Ann Thorac Surg
71: S433–S436, 2001.
23.Matsuura JH, Black KS, Levitt AB, et al: Cellular remodeling of depopulated bovine ureter used as an arteriovenous graft in the canine model. J Am Coll Surg
198: 778–783, 2004.
24.Sharp MA, Phillips D, Roberts I, Hands L: A cautionary case: the synergraft vascular prosthesis. Eur J Vasc Endovasc Surg
27: 42–44, 2004.