Biomechanical Characterization of a Micro/Macroporous Polycaprolactone Tissue Integrating Vascular Graft

Biomechanical Characterization of a Micro/Macroporous

Polycaprolactone Tissue Integrating Vascular Graft

YIWEI WANG,1,6 JENNY LAM,2,7 BUFA ZHANG,2 PAUL E. TOMLINS,2 XIONGWEI LI,3 OYA ALPAR,3

DAVID F. WERTHEIM,4 ALLAN S. JONES,5 and ALLAN G. A. COOMBES1,8

1School of Pharmacy and Chemistry, Kingston University, London, Kingston upon Thames, Surrey KT1 2EE, UK; 2NationalPhysical Laboratory, Hampton Road, Teddington, Middlesex TW11 0LW, UK; 3School of Pharmacy, University of London,

Brunswick Square, London WC1N 1AX, UK; 4Faculty of Computing, Information Systems and Mathematics, KingstonUniversity, London, Kingston upon Thames, Surrey KT1 2EE, UK; 5Australia Key Center for Microscopy and Microanalysis,University of Sydney, Sydney, NSW 2006, Australia; 6Burns Unit, ANZAC Research Institute, Concord Repatriation GeneralHospital, Hospital Road, Sydney, NSW 2139, Australia; 7Department of Pharmacology and Pharmacy, The University of HongKong, 21 Sassoon Road, Pokfulam, Hong Kong; and 8School of Pharmacy, The University of Queensland, St Lucia, Brisbane,

QLD 4072, Australia

(Received 8 April 2010; accepted 5 July 2010; published online 20 July 2010)

Associate Editor Jay Humphrey oversaw the review of this article.

Abstract—The objective of the present study was to charac-terize the short-term biomechanical properties of cast micro/macroporous poly(caprolactone) (PCL) tubes intended forapplication as tissue integrating blood vessel substitutes.Micro/macroporous PCL vascular grafts (5.5 mm internaldiameter, 7.5 mm external diameter) with defined macroporestructures were produced by rapidly cooling PCL solutionscontaining dispersed gelatin particles in dry ice, followed bysolvent and gelatin extraction. A Bose-Enduratec Bio-Dynamic chamber configured for cardiovascular applicationswas used to measure the diametrical stability (dilation) oftubular samples under hydrodynamic flow conditions at37 �C. Microporous PCL tubes withstood the hydrodynamicstresses induced by short, 2-min duration flow rates up to1000 mL/min, which resulted in estimated internal pressuresin excess of arterial pressure (80–130 mmHg). Micro/macro-porous PCL tubes having a maximum macroporosity of 23%accommodated the hydrodynamic stresses generated by shortduration, flow rates up to 1000 mL/min, which resulted inestimated internal pressures similar to venous pressure(30 mmHg).The dilation of microporous PCL tubes undershort, (5 min) pulsatile flow conditions (1 Hz) increased from10 to 100 lm with increasing mean flow rate from 50 to500 mL/min. Both microporous and macroporous tubesexhibited a burst strength higher than 900 mmHg underhydrostatic fluid pressure, which is in excess of arterialpressure (80–130 mmHg) by a factor of approximately 7.Quantitative analysis of the macropore structure was per-formed using micro-computed tomography for correlationwith mechanical properties and cell growth rates. Mousefibroblasts efficiently colonized the external surface of

macroporous PCL materials over 8 days in cell culture andcell numbers were higher by a factor of two compared withmicroporous PCL. These findings demonstrate that micro/macroporous PCL tubes designed for vascular tissue engi-neering can accommodate the hydrodynamic stresses gener-ated by short duration, simulated blood flow conditions andexhibit good potential for integration with host tissue.

Keywords—Polycaprolactone, Porous-walled tubes, Scaf-

folds, Vascular grafts, Biomechanical properties, X-ray

microcomputed tomography lCT, Fibroblast.

INTRODUCTION

Advanced stages of vascular disease such asobstructive atherosclerosis or aneurysms generallynecessitate replacement of blood vessels with vascularprostheses. However, autografts (e.g., saphenous veinsfor lower extremity bypass procedures) are unsuitablein around 10–30% of patients.11,21,32 Synthetic grafts(>10-mm diameter) produced from Dacron polyesterfibers or expanded poly(tetrafluoroethylene) (ePTFE)have performed successfully in large vessel replacement(e.g., the aortic/iliac arteries) which are characterizedby high blood flow and low flow resistance, resulting in85–95% patency at 5 years. However, they are notrecommended for small blood vessel replacement (e.g.,popliteal or tibial) because of poor patency rates.Saphenous veins are preferred. A layer of fibrin andfibrous tissue, respectively, forms on the intimal andouter surface of synthetic grafts soon after implanta-tion. Formation of a pseudointima may be followed at

Address correspondence to Allan G. A. Coombes, School of

Pharmacy, The University of Queensland, St Lucia, Brisbane,

QLD 4072, Australia. Electronic mail: a.coombes@pharmacy.

uq.edu.au

Cardiovascular Engineering and Technology, Vol. 1, No. 3, September 2010 (� 2010) pp. 202–215

DOI: 10.1007/s13239-010-0019-1

1869-408X/10/0900-0202/0 � 2010 Biomedical Engineering Society

202

a later stage by pseudointimal hyperplasia (a conditionarising from uncontrolled growth of randomly ori-ented smooth muscle cells (SMCs) embedded inextracellular matrix (ECM) and is a major cause ofgraft failure and clinical concern. Compliance mis-match between the implant and host artery, resulting indisrupted blood flow conditions at the anastomosis orgraft/artery junction, has long been implicated inprosthesis failure arising from anastomotic pseudoin-timal hyperplasia.11 Thus, a long standing design cri-terion for synthetic grafts is that they should display adynamic mechanical response in vivo similar to that ofthe host blood vessel. Knitted Dacron prostheses, forexample, exhibit a stretch characteristic while crimpingimparts further elasticity to the structure. Tissue inte-gration with the external graft surface is encouraged toprovide natural reinforcement of the graft againstcompression. Production of ePTFE grafts involvesmechanical stretching of the material, resulting in afibrillar structure which favors integration with fibro-vascular tissue.

A large number of design and production strategieshave been applied over many years to improve theperformance of synthetic vascular grafts includingendothelial cell (EC) seeding,1,25 modification of theluminal surface and controlled delivery of bioactives toimprove thromboresistance,17,20 the use of bioresorb-able scaffolds to support growth of new blood ves-sels8,9 and the production of totally biologic orbiomimetic structures for transplantation.12 However,the continued surgical reliance on Dacron textileprostheses and ePTFE provides a stark reminder of thelack of progress in developing alternatives. Vasculargraft production from bioresorbable fibers such aspoly(glycolic acid) (PGA)8 and poly(L-lactic acid)(PLA)14 aims to provide a temporary framework forregenerating tissue to form a new blood vessel. After4 weeks post-implantation, the inner capsule formedon Dacron grafts generally comprises fibrin coagulumand few cells, whereas, woven PGA grafts have beenshown to produce an inner capsule comprising a con-fluent layer of ECs, smooth muscle-like myofibroblastsand dense collagen fibers.8 However, aneurysmaldilation is a major problem for these grafts, due torapid resorption of the PGA scaffold in 3 months, i.e.,before sufficient tissue ingrowth and strength devel-opment has occurred to resist hemodynamic pressures.Early attempts to tissue engineer vascular grafts byseeding vascular cells on biodegradable scaffolds wereesthetically promising but functionally lacking. Shin-oka et al.22 seeded ECs on PGA scaffolds. Scaffoldresorption occurred in 11 weeks in vivo, leaving ECslining the luminal surface of a neoartery and a devel-oped ECM. Niklason et al.18 cultured SMCs on a PGAscaffold subjected to pulsatile stress for 8 weeks

followed by seeding of ECs on the luminal surface for3 days. Although impressive morphological andmechanical properties were measured, animal studiesrevealed decreased blood flow through the graft4 weeks post-implantation. Resorbable textile graftscontinue to be a favoured design option in vasculartissue engineering. Matsumura et al.14 produced dura-ble grafts (24 mm in diameter) for clinical trials inhumans by seeding CD34-positive cells (isolated frombone marrow cell aspirates) onto woven poly(L-lactide)fabric tubes on the day of surgery. An absence ofthrombogenic complications, stenosis and graftobstructionwas reported. L’Heureux et al.12 produced a‘‘totally biologic’’ vascular graft exclusively from cul-tured cells to form anEC luminal layer, SMCmedia andfibroblast adventitia but intramural blood infiltrationresulted in short-term in vivo studies. Daunting techni-cal problems associated with 3-D tissue architecture,cell–cell interaction and ECM organization, combinedwith complex production and concerns over long-termpatency due to graft degeneration are expected toseverely restrict progress toward successful clinicalapplication of totally biologic substitutes for bloodvessels.

The objective of the present study was to charac-terize the short-term biomechanical properties of castmicro/macroporous poly(caprolactone) (PCL) tubesintended for application as tissue integrating bloodvessel substitutes. The favorable biocompatibility ofPCL with soft and hard tissue has led to widespreadinvestigations of the polymer for tissue engineering,5,29

with several describing vascular graft production.13,30

The tube flexibility and degree of elasticity resultingfrom PCL’s tensile modulus (E) 0.4 GPa, tensilestrength 29 MPa, failure extension (>700%)16 alignwith the low compliance of natural blood vessels toreduce the incidence of anastomotic hyperplasia. Thelow biodegradation rate of PCL (>4 years resorptiontime in vitro4) imparts long-term resistance to hemo-dynamic pressure and aneurysm formation, in contrastto bioresorbable polymers such as PGA and polydi-oxanone (PDS). The microporous character of the tubeprovides a route for nutrient and metabolite transport,while the macroporous structure is intended initially topromote rapid integration of the graft with surround-ing soft tissue to provide reinforcement againsthemodynamic stress. The persistence of the scaffoldsupports gradual ingrowth of fibrovascular tissue tobuild a replacement blood vessel over time in vivo.Tube dilation and burst strength were measured undershort-term hydrodynamic and hydrostatic loadingconditions, respectively. Macroporosity was quantita-tively defined using X-ray micro-computed tomogra-phy (l-CT) image analysis. The potential for graftintegration with host fibrovascular tissue was assessed

Biomechanical Characterization of a Micro/Macroporous Polycaprolactone Tissue Integrating Vascular Graft 203

by monitoring the interaction of fibroblasts in cellculture with tubular PCL scaffolds featuring definedmacropore size ranges.

MATERIALS AND METHODS

Materials

PCL (Mw = 115 kDa, Capa 650) was obtained fromSolvay Interox, Warrington, UK. Gelatin powder(type B, bloom 125), phosphate buffered saline (PBS)and bicinchoninic acid (BCA) reagents were purchasedfrom Sigma-Aldrich Chemicals. Acetone and methanolwere purchased from Fisher Scientific, UK.

Production of Microporous, and Micro/MacroporousPCL Vascular Grafts

Micro/macroporous PCL vascular grafts were pro-duced by a rapid cooling technique. PCL (1.7 g) wasdissolved in 10 mL acetone at approximately 50 �C toproduce a 17% w/v solution. Gelatin powder (insolu-ble in acetone) was sieved to obtain a particle sizerange of 90–125 lm and dispersed in 17% w/v PCLsolutions to produce three protein concentrations of29, 38, and 44% w/w, respectively. The suspensionswere transferred into a mold comprising a 3 mLpolypropylene (PP) syringe body with a centrally lo-cated 1 mL PP syringe body and rapidly cooled in dryice for 3 min. The tubular castings obtained (nominalinternal diameter, 5.5 mm external diameter, 7.5 mm,length 60 mm) after crystallization and hardening ofthe polymer, were removed from the mold and im-mersed in methanol (50 mL) for 24 h to extract theacetone. Any acetone/methanol remaining in thematrices was allowed to evaporate in air under ambientconditions. Micro/macroporous grafts were obtainedby extracting the gelatin phase in PBS at 37 �C over21 days to produce samples with theoretical macrop-orosities of 13, 19, and 23%, respectively. The theo-retical macroporosity was calculated using the weightof PCL and gelatin in the casting, a density of 0.39 g/cm3 for the microporous PCL phase2 and 1 g/cm3 forgelatin. Microporous PCL grafts were also producedfrom 17% w/v PCL solution without dispersed gelatinparticles, using the above method.

Mechanical Evaluation

Measurement of the Changes in Diameter of PCLVascular Grafts Under a Constant Flow Rateof 300 mL/min

A Bose-Enduratec BioDynamic chamber configuredfor cardiovascular applications (BOSE, Minnesota,

USA, Fig. 1) was used to mimic human blood flowrates under steady state and pulsatile flow conditionsat 37 �C. For the former study distilled water wascontinuously circulated at a prescribed flow rate from areservoir through a mixing chamber before passingthrough the sample tube and back to the reservoir.

A PCL tube was mounted on 6-mm diameter Luerfittings (Kynar) and held vertically in the center of thesample chamber. The chamber and sample tube werefilled with distilled water at 37 �C and a constant flowrate of 300 mL/min was maintained inside the tube for1 h to simulate blood flow in the vasculature.26 Sam-ples of microporous PCL tubes and micro/macropo-rous PCL tubes were tested in triplicate. Changes inouter diameter were measured over a time interval of2 min at intervals of 10 measurements per second usinga laser scanning micrometer (Mitutoya LSM 301)configured as a simple shadowgraph detector. Datawere captured at a sensitivity of ±0.1 lm and at 16 bitresolution and all data were logged using a Labview8.0 based programme (National Instruments, US).

Measurement of PCL Tube Dilation Under Increasing,Continuous Flow Rates

The change in outer diameter (dilation) of PCLtubes under increasing flow rates was investigated toassess their short-term recovery characteristics fol-lowing exposure to hydrodynamic stress. MicroporousPCL tubes and micro/macroporous PCL tubes weretested using the Bose-Enduratec biodynamic chamberdescribed above. The flow rate of distilled waterthrough the sample tubes was gradually increased by100 mL/min increments at 2 min intervals to a maxi-mum of 1000 mL/min. After 2 min at 1000 mL/min,the flow rate was reduced to a flow rate of 50 mL/minand the test was repeated twice. Changes in the outer

FIGURE 1. The Bose-Enduratec biodynamic chamber.

WANG et al.204

diameter of the tubes were recorded using a MitutoyaLSM 301 laser scanning micrometer at a rate of10 scans/s.

Measurement of PCL Tube Dilation Under PulsatileFlow Conditions

The Bose-Enduratec biodynamic chamber can beconfigured to provide pulsatile flow conditions throughthe action of a piston, which modulates fluid flowthrough the mixing chamber (Fig. 1). A sinusoidalwaveform and pulse frequency of 1 Hz were applied.Values of the minimum and maximum flow ratesdefined the amplitude of the dynamic flow rate and amean value. In this case the minimum flow rate was setto zero. A range of mean flow rates (50–600 mL/min)of distilled water at 37 �C was investigated and flowwas maintained for 5 min at each value. The outerdiameter of the sample tube was measured over a 50 stime interval mid-way through the test using aMitutoya LSM 301 laser scanning micrometer at a rateof 10 scans/s.

Determination of the Burst Strength of PCL Tubes

The burst strength of PCL tubes was investigatedunder hydrostatic conditions using the test set-upshown in Fig. 2. Samples were filled with distilled

water and immersed horizontally in a water bathmaintained at 37 �C. A short length of silicone tubewas used to couple one end of the PCL sample tube toa compressed air supply via a clamp. A second sectionwas used to seal the other end of the PCL tube. Inkwas added to the water in the sample tube to visuallyidentify any incidences of tube cracking prior to sam-ple failure. A pressure of 102 kPa was initially appliedto the sample following temperature equilibration andincreased at a rate of 50–100 Pa/s until the burst pointwas reached. Samples of microporous PCL tubes andmicro/macroporous PCL tubes were tested in triplicateto determine the mean burst strength. The change intube diameter with increasing pressure was measuredusing a Mitutoyo LSM 301 laser scanning micrometer,at a rate of 10 scans/s.

l-CT Analysis of the Internal Pore Structure

The 3D pore structure of micro/macroporousPCL vascular graft material was analyzed by l-CT asdescribed previously.27 In brief, samples (approxi-mately 2 9 29 10 mm in length) were cut longitudi-nally from PCL tubes and analyzed using a Skyscan1072 (Skyscan, Aartselaar, Belgium) desktop X-ray CTscanner at 15 lm voxel resolution (509 magnification),X-ray tube current 173 lA and voltage 30 kV.

FIGURE 2. Schematic representation of the test facility used for burst testing of PCL vascular grafts.


Specimens were mounted vertically and rotatedthrough 360� around the long axis. Absorptionimages were recorded every 0.225� of rotation andused in standard cone-beam reconstruction softwareto generate a series of 1024 8 bit axial slices. Three-dimensional reconstruction of the internal pore mor-phology was carried out using these axial bitmapimages and analyzed by VG Studio Max 1.2 software(Volume Graphics GmbH, Heidelberg, Germany).Analysis of transverse 2D l-CT ‘‘shadow’’ imagesobtained at 1000 lm intervals along the sample longaxis provided a quantitative estimation of pore sizedistribution in the PCL tubes. The variation of‘‘equivalent pore diameter’’ in each image slice wasassessed using MINITAB (Minitab Inc., USA) andcombined to provide the frequency distribution ofequivalent pore diameters throughout the material.The analysis was confined to macropores withequivalent diameter larger than 16.6 lm due to imageresolution constraints.

Growth of Fibroblasts on PCL Vascular Graft MaterialHaving Various Pore Size Ranges

Swiss 3T3 mouse fibroblasts (purchased from theEuropean Collection of Animal Cell cultures(ECACC) were seeded at a density of 10 9 103 cells/cm2 in 48-well tissue culture plastic plates (TCP) con-taining microporous PCL vascular graft material andmicro/macroporous material (13% theoreticalmacroporosity). The latter materials featured specificmacropore size ranges of 45–90, 90–125, and125–250 lm, respectively. Samples in the form of disks(1 cm in diameter) were cut from tubular castings andsterilized under UV light in a cell culture hood over-night (12 h). Samples were seeded in triplicate to gainmeasurements of cell numbers on each type of materialat various time points up to 8 days. Cells were alsoseeded on TCP as a control. Plates were incubated at37 �C and 5% CO2 in DMEM culture medium (10%v/v fetal bovine serum (FBS), 1% v/v L-glutamine(2 mM/L final conc.) and 1% v/v antibiotic solution(100 IU/mL penicillin and 100 lg/mL streptomycin).The medium was replaced with fresh DMEM every2 days. Growth of cells on the materials was allowedfor 8 days and cell numbers were counted at day 1, 5,and 8. Each sample with attached fibroblasts was thenwashed using sterile PBS to remove residual culturemedium and cells were detached by adding 0.25% w/vtrypsin solution in EDTA (0.3 mL) to each sample,followed by incubating at 37 �C for 5 min. Trypsin wasinactivated by adding 0.5 mL culture medium to eachwell. The number of cells detached from the matriceswas counted using a Weber haemocytometer.

Scanning Electron Microscopy

The fracture surface of PCL tubes following bursttesting was examined using a ZEISS EVD 50(Germany) scanning electron microscope (SEM).Samples were attached to aluminum SEM stubs usingcarbon tabs and sputter coated with platinum using anEdwards E30GA automatic mounting press prior toexamination in the SEM at a voltage of 10 kV. Themorphology of Swiss 3T3 fibroblasts attached at 1 and5 days on PCL graft materials having various pore sizeranges was examined using a Philips XL30 SEM (FEI,Eindhoven, the Netherlands). Samples with attachedfibroblasts were removed from the 48-well TCP plates,transferred to 10 mL centrifuge tubes and residualculture medium was replaced by sterile PBS (pH 7.4).After rinsing three times with PBS, the fibroblasts at-tached to the surface of the materials were fixed byaddition of 2.5% glutaraldehyde in 0.1 M sodiumcacodylate buffer for 30 min (first fix). Samples werethen rinsed with PBS (3 9 5 min) before the second fixusing 2% osmium tetroxide in 0.2 M sodium caco-dylate buffer for 1 h. Cells attached to the PCL sam-ples were washed with sodium cacodylate buffer anddehydrated using a series of ethanol dilutions(50–100%) for 5 min, twice per dilution. Specimenswere finally dried in a critical point dryer, mounted onaluminum sample stubs and sputter coated with plat-inum prior to examination in the SEM at a voltage of15 kV.

RESULTS AND DISCUSSION

Exposure of PCL solutions containing dispersedgelatin particles to dry ice induces rapid crystallizationof the polymer which is advantageous for maintaininga uniform distribution of particles. Extraction of ace-tone from the hardened material, using methanol,followed by extraction of gelatin produces a soft-textured, micro/macroporous, highly flexible materialon drying which is free of large-scale cracks and voidson both the surface and in the interior. Gelatinextraction, post-implantation offers opportunities forco-delivery of bioactives such as antibacterials andgrowth factors to improve the function of vasculargrafts. The macroporosity of cast tubes was limited to23% in these investigations since the inclusion of highloadings of gelatin particles resulted in deterioration ofthe tube strength. The microporosity of the PCL phase(5–15 lm pore size range) is expected to be similar toprecipitation cast PCL matrices, namely 68–77%.2

Vascular graft designs are required to withstandlocal hemodynamic conditions of blood flow rate,pressure and pulse frequency. Thick-walled, large

WANG et al.206

arteries such as the aorta (10–30 mm id, 2–3 mm wallthickness) and main distributing branches must with-stand pulsating 80–130 mmHg pressures. Smallerarteries (5 mm id, 1 mm wall thickness) operate atsteadier blood pressures in the 70–90 mmHg range,whereas the venous system is exposed to relativelyconstant lower pressures which can vary from30 mmHg to around zero depending on the anatomicallocation.21 Sufficiently compliant tubes dilate by anamount dependent on the pressure of the fluid, gener-ating a hoop or circumferential strain in the tube wall.7

The pulsatile nature of the circulatory system thusproduces a dynamic hoop stress and hence strain, themagnitude of which depends on the tube dimensions, itsviscoelastic properties and location within the body.Blood flow also generates a shear stress or tangentialdrag force at the wall of the blood vessel, whose mag-nitude is directly proportional to the fluid viscosity andflow rate and inversely proportional to the cube of thetube radius.19 Cheng et al.3 measured a shear stress of3.4 and 2.3 dynes/cm2, respectively, at the supraceliacand infrarenal regions of a human abdominal aorta.

In general, fluid flow through microporous PCLtubes, established using the BioDynamic chamber in-duced dilation that increased with the duration of flow.The outer diameter of microporous PCL tubesincreased by (approximately 10 lm) in a period of 1 hunder a constant flow rate of 300 mL/min at 37 �C(Fig. 3a). Measurements of tube diameter at a partic-ular time point are affected by ‘‘noise’’ associated withthe high sensitivity of the laser micrometer. The sig-nificant fluctuation in diameter measured at 30 minmay be caused, by transient errors in the micrometerresponse or air bubbles on the tube surface. Gelatin-loaded tubes (corresponding to 13, 19, and 23%macroporous samples) exhibited an initial reduction indiameter of 100–200 lm during 15–30 min testing at300 mL/min, before remaining constant. This behavioris attributed to the sensitivity of the laser scanningmicrometer and its response to dissolution of the(90–125 lm) gelatin particles protruding from the tubesurface. The outer diameter of macroporous PCLtubes with pore size range of 90–125 lm and theoret-ical macroporosity of 13, 19, and 23% remained fairly

FIGURE 3. The changes in outer diameter of (a) microporous PCL tube, (b–d) 13, 19, and 23% macroporous PCL tubes over 1 h at37 �C under a constant flow rate of 300 mL/min. Macropore size range 5 90–125 lm.


constant for 1 h under a flow rate of 300 mL/min at37 �C (Fig. 3b–d). The oscillation of outer diameter atthe start of testing (Fig. 3b) may be caused by air beingforced out of the porous tube walls and forming bub-bles that adhere to the outer surface. The absence ofsignificant dilation suggests that the macroporousstructure can easily resist the hydrodynamic stressesassociated with a 300 mL/min fluid flow rate overshort time scales.

The outer diameter of microporous PCL tubes andmacroporous PCL tubes was also measured underincreasing flow rates from 50 to 1000 mL/min. Threerepeat tests were performed on each sample to generateinformation on the short-term recovery characteristicsof tubular samples subjected to internal fluid flow. Theouter diameter of microporous PCL tubes increasedinitially by approximately 40 lm at a flow rate of1000 mL/min (Fig. 4a). On reducing the flow rate to50 mL/min the outer diameter was found to haveincreased by at least 10 lm compared with the startingdiameter. Thus, the microporous PCL tube does notimmediately recover its original dimensions after beingsubjected to short duration, high flow rates. In thesecond run, the outer tube diameter first decreased and

then gradually increased with increasing flow rate,indicating an ongoing recovery process. The diameterof the PCL tube at each flow rate was similar in thesecond and third test, indicating that permanentexpansion (‘‘set’’) of PCL tubes is produced initiallyunder high flow rates, but the viscoelastic recoverycharacteristics of the ‘‘stress modified’’ or conditionedmaterial limits further deformation.

Under continuous increasing flow conditions(Fig. 4b–d) the outer diameter of 13% macroporousPCL tubes (90–125 lm pore size) increased rapidlywith flow rate by approximately 5 lm and then tendedto plateau at flow rates of 400–1000 mL/min. Inaddition, the sample tube did not fully recover itsoriginal diameter after being subjected to high flowrates; approximately 5 lm dilation remained. Similarbehavior to microporous PCL tubes (Fig. 4a) wasobserved for 19 and 23% macroporous PCL tubes, inthat increasing dilation occurred with increasing flowrate during initial testing resulting in a small residualdeformation (or ‘‘set’’) of approximately 5–10 lm(Fig. 4c, d). Good recovery of tube diameter occurredfor 19 and 23% macroporous samples after the secondand third runs from a maximum dilation of around 15

FIGURE 4. The changes in outer diameter of (a) microporous PCL tube, (b–d) 13, 19, and 23% macroporous PCL tubes (90–125 lmpore size) under continuous flow at 37 �C (1, 2, and 3 are repeat measurements on the same sample).

WANG et al.208

and 30 lm, respectively. This behavior demonstratesthat macroporous PCL tubes are also initially condi-tioned during hydrodynamic loading and can subse-quently accommodate the hoop strain induced bycontinuous flow conditions in the short-term withoutsignificant permanent deformation. Pre-stressing ofPCL tubes could be advantageous to introduce a per-manent ‘‘set’’ condition prior to implantation. Theextent of dilation increases with increasing macropo-rosity of the matrix reflecting the reduction in thepolymer phase and the consequent increase in tubecompliance.

Blood flow rates of around 320 mL/min are expe-rienced in the celiac, superior mesenteric and renalarteries during mild exercise conditions.26 A compari-son of the changes in tube dilation during flow testing(Fig. 4) and the incremental changes in internal pres-sure with tube dilation during burst testing, indicatedthat microporous PCL tubes were subjected to pres-sures of around 450 mmHg under flow conditions of1000 mL/min. In contrast, the micro/macroporoustubes experienced pressures <40 mmHg at this flowrate which are closer to high venous pressures(30 mmHg). The macroporous PCL tubes accommo-date the hoop stress and strain by deformation of thepore structure. However, interstitial fluid flow (leak-age) through the porous tube wall may provide apressure relief mechanism through hydraulic perme-ability and viscous or frictional drag effects.15,35 Thereduced dilation of macroporous tubes compared withmicroporous tubes (Figs. 3 and 4) is consistent with abehavior of increasing pressure relief with increasingpermeability of the tube wall. At the maximum flowrate investigated of 1000 mL/min, a similar dilation of30 lm was obtained for microporous tubes and gela-tin-loaded tubes corresponding to 13 and 19% mac-roprous samples. Dilation was increased to 60 lm forgelatin-loaded tubes compared with 30 lm for thecorresponding 23% macroporous samples. Thus, fill-ing of the macropores with gelatin tends to increasetube dilation providing a further indication thatrestricted fluid flow through the tube wall maintainshigh internal pressures.

Direct evidence for transmural fluid flow is pre-sented in our recent study of Darcy permeabilitycoefficients for microporous PCL tubes28 which pro-vided information relevant to nutrient, metabolite andgrowth factor transport through the tube wall. Thetransmural flow rates of water, PBS and glucosesolution for tubes of similar dimensions to those de-scribed in the manuscript were similar and equate to apermeation rate of around 2.6 mL/s at maximumarterial blood pressure of 130 mmHg. The permeationrate of albumin solution was significantly lower(approximately 1.1 mL/s) indicating pore sealing by

protein adsorption. The macroporous PCL tubes areexpected to show higher leakage rates than micropo-rous samples, which could be reduced by implantinggelatin-loaded tubes and allowing gelatin extraction tooccur post-implantation. In vitro testing indicates thatthe gelatin phase would be gradually released over aperiod of 3 weeks.27 The permeability coefficient of theaorta (rabbit) is around 1.4 9 10�14 cm2, therebylimiting transmural interstitial fluid flow in the arterywall to levels which are six orders of magnitude lowerthan blood flow.23 No blood leakage is a basicrequirement of vascular prostheses and is achieved inpractice by sealing textile designs with collagen, albu-min or gelatin.24 Knitted Dacron grafts, for example,exhibit a permeability of 1500 mL cm�2 min�1 butcollagen coatings have been found to reduce the waterpermeability to 6–9 mL cm�2 min �1 rendering thegraft impervious to blood.10 Thus, micro/macroporousPCL grafts could be sealed with gelatin/collagen aspracticed for porous, knitted Dacron prostheses tominimize blood leakage following implantation. Alayer of fibrous tissue is known to cover the outersurface of synthetic vascular grafts soon after implan-tation. Tubular foam scaffolds of poly(lactide-co-gly-colide), for example, are encapsulated by fibrous tissue1 week post-implantation in Adult Lewis rats.6 Thus,rapid integration of the external PCL graft surfacewith host tissue is anticipated to provide a permanentsealant function against blood leakage.

The changes in outer diameter of a microporousPCL tube under short-term, pulsatile flow conditionsat 37 �C are shown in Fig. 5. Tube dilation increasedsignificantly with flow rate (Table 1) and by a factor of2.5 compared with continuous flow conditions. Theouter tube diameter increased by approximately5–10 lm at a low mean flow rate of 50 mL/min and by100 lm at 500 mL/min. In addition the outer tubediameter was found to have increased by approxi-mately 40 lm at the end of the test indicating incom-plete recovery as observed for continuous flowconditions (Fig. 4). Stress conditioning and reinforce-ment of the graft by integration with host fibrovasculartissue are expected to reduce tube dilation under pul-satile flow conditions in vivo.

The burst strength of microporous PCL tubes andmacroporous PCL tubes (90–125 lm pore size range)was investigated under hydrostatic conditions at37 �C. Two distinct responses were observed in thepressure vs. time profile (Fig. 6). At first, the pressureincreased gradually then dropped sharply withoutvisible failure of the tube. In the second stage, a sharpincrease of pressure occurred until the burst point ofthe tube was reached when longitudinal cracking wasapparent. The first point of pressure drop and the burstpoint of microporous PCL tubes was recorded at


132 kPa (990 mmHg) and 149 kPa (1118 mmHg),respectively (Table 2). The outer tube diameterincreased gradually by approximately 20 lm (follow-ing the initial pressure application of 102 kPa) until thefirst point of pressure drop and then remained con-stant. The longitudinal cracking observed during bursttesting of porous PCL tubes indicates that hoopstresses are responsible for tube failure.7 The sharpdrop in pressure recorded during burst testing of PCLtubes, may indicate failure of the luminal surface, priorto crack propagation through the tube wall. Thiscapacity for ‘‘pressure relief’’ could be advantageous inavoiding catastrophic failure of implanted tubes.

Macroporous PCL tubes exhibited decreased burststrength (900–950 mmHg) and a lower first pointpressure reading compared with microporous PCL

tubes (Table 2) which may be explained by the reduc-tion in content and continuity of the PCL phase in thetube wall. Tube dilation at burst was significantlyhigher than microporous samples, at around 400, 500,and 150 lm for 13, 19, and 23% macroporous sam-ples, respectively. The burst strength may be improvedif necessary by controlling factors, such as the wallthickness or by incorporating an outer reinforcingsleeve in the graft design. Natural reinforcement ofmacroporous PCL grafts is expected in vivo followingintegration with host tissue. The burst strength ofmicroporous and micro/macroporous PCL tubesexceeds by a factor of approximately 7, the bloodpressure of 80–130 mmHg experienced by large arter-ies (10–30 mm diameter).21 In comparison, the rupturestrength of saphenous veins and ‘‘totally biologic’’

FIGURE 5. The changes in outer diameter of microporous PCL tube under pulsatile flow conditions at 37 �C (Exp 1, 4, 8, and 10 inTable 1).

TABLE 1. The effect of pulsatile flow conditions on thedilation of microporous PCL tubes.

Exp.

no.

Max dynamic

flow rate (mL/min)

Mean flow

rate (mL/min)

Change in

outer tube diameter (lm)

1 100 50 10

2 200 100 20

3 300 150 30

4 400 200 40

5 500 250 50

6 600 300 55

7 700 350 75

8 800 400 80

9 900 450 90

10 1000 500 100

Pulse frequency = 1 Hz, minimum dynamic flow rate = 0 mL/min,

PCL tube length = 60 mm, outer tube diameter = 7.7 mm.

FIGURE 6. The changes in outer tube diameter and internalpressure during burst testing of microporous PCL tubes.

WANG et al.210

tissue engineered grafts has been recorded as 1680 and2150 mmHg, respectively.1,12,25 Lee et al.13 obtainedburst strengths of around 500 mmHg for PCL/colla-gen composite vascular grafts produced by electro-spinning, but the use of an impermeable latex tube toapply internal pressure in the latter study does notpermit assignment of a realistic burst strength to thesestructures.

The fracture surface of microporous PCL tubesfollowing burst testing displayed a rough texture withirregular pore shape and size of 5–20 lm (Fig. 7a, b).The PCL fibrils visible in the fracture area (arrowed)are probably formed by deformation of the PCL phaseat the burst point. SEM analysis of macroporous PCLtubes revealed the added presence of macropores of

size 90–125 lm (Fig. 7c, d) formed by extraction ofgelatin particles. Macroporous tubes exhibited a lessfibrous fracture morphology than the microporoussamples and these areas tended to be more localized(arrowed in Fig. 7c). No significant differences wereapparent between the fracture surfaces of PCL tubeswith 13, 19, and 23% macroporosity.

The above findings of increased tube dilation anddecreased burst strength with increasing macroporos-ity of the matrix underline the importance of definingthe scaffold pore structure to ensure reproducibility ofproperties. l-CT analysis of the internal 3-D structureof macroporous PCL grafts, clearly defined themicroporous PCL phase and the macroporouscomponent (Fig. 8a, b). The shape and size of the

TABLE 2. Burst strength of microporous PCL and micro/macroporous PCL tubes.

PCL tube First point of pressure drop (kPa) (mmHg) Burst point kPa (mmHg)

Microporous 131.7 ± 4.8 (988) 148.7 ± 12.4 (1115)

13% Macroporous 114.6 ± 3.9 (860) 126.6 ± 5.4 (950)

19% Macroporous 115.6 ± 1.1 (867) 125.3 ± 5.7 (940)

23% Macroporous 115.1 ± 2.5 (863) 120.6 ± 6.6 (905)

FIGURE 7. Scanning electron micrographs of the fracture surface of PCL tubes following burst testing (a, b) microporous PCL.Arrow in (b) indicates fibrils formed by deformation of PCL (c) 13% macroporous PCL. Arrow indicates region of fibrous mor-phology in fracture surface (d) 23% macroporous PCL. Arrow indicates macropores exposed in fracture surface.


macropores corresponds closely with the particles usedfor tube production. The high pore density revealed inthe 23% macroporous materials produces a network ofinterconnected macropores in certain areas (Fig. 8c)due to particle contact and subsequent particle disso-lution, which provide pathways for cell migration intothe graft interior. This feature is expected to promoteintegration of the implanted graft with host tissue. Thefrequency distribution of macropore size (in terms ofequivalent pore diameter) is shown in Fig. 8d and thecorresponding summary of macropore size for thematerials investigated is presented in Table 3. Spatialdifferences in microporosity (5–15 lm pore size) across

the tube wall were not obtained since the resolution ofthe Skyscan 1072 micro-CT facility was limited to aresolution of 15 lm.

Cell culture experiments revealed that the number offibroblasts attached to macroporous PCL matrices wassimilar to TCP on day 1 but lower than TCP by afactor of 2 and 3 at day 5 and 8, respectively (Fig. 9).Enhanced cell attachment and proliferation on TCPhas been recorded in many studies and is due to themodified surface chemistry developed specifically forcell culture. SEM analysis of seeded fibroblasts at day1 on macroporous PCL samples (Fig. 10b) revealedlarge numbers of rounded cells (approximately 10 lm

FIGURE 8. PCL vascular graft material (23% macroporosity, 90–125 lm pore size). Internal microtomographs of (a) microporouspolymer phase (b) macroporous structure (c) interconnected macropores (d) frequency distribution of equivalent pore diameter.

TABLE 3. Summary of micro-CT analysis of macropore size in PCL vascular graft materials (selected pores having an equivalentdiameter >16.6 lm).

Theoretical

macroporosity (%)

Sieved particle

size range (lm)

Median

(inter-quartile range)

equivalent pore diameter (lm)

Total range

equivalent pore

diameter (lm)

90% Pores

less than (lm)

10% Pores less

than (lm)

Fractional

pore area (%)

13 90–125 34.2 (25.2–44.4) 16.8–115.9 56.6 19.8 6

19 90–125 52.1 (40.3–72.1) 17.1–169.5 88.5 25.0 11

23 90–125 23.7 (19.4–38.3) 16.6–179.3 67.8 17.8 18

WANG et al.212

in size) in contact with the surface. Long processextensions linking the cells to the underlying matrixand to neighboring cells are also apparent. Pronouncedflattening and spreading of cells was observed at day 3,obscuring the underlying polymer surface, with cellprocesses connecting the flattened cell bodies. Thisbehavior indicates strong cell adhesion with theunderlying substrate via focal contacts and provides ameasure of the biocompatibility of the polymer. Simi-lar behavior was observed previously for 3T3 mousefibroblasts cultured on PCL fibers.31

Cell numbers were generally higher by a factor oftwo on macroporous PCL at day 5 and 8 comparedwith microporous material. However, the macropo-rous component is not anticipated to influence thenumber of attached cells due to the low macroporedensity at the material surface (Fig. 10a) and the lim-ited time available for any cell migration into thesample interior. Instead, residual gelatin moleculesbound to the PCL scaffold during particle solubiliza-tion are expected to be more influential in enhancingfibroblast proliferation by enabling integrin-mediatedbinding to RGD cell adhesion sequences present in thedenatured collagen molecule. Gelatin coating of PCLfibers has previously been shown to improve fibroblastproliferation by a factor of 2 after 48 h in cell culture.31

The data in Fig. 9 also indicates that a macropore sizeof 90–125 lm is advantageous for fibroblast growth onmacroporous PCL materials. These findings arebroadly in line with those of Yang et al.34 who foundthat the number of human skin fibroblasts cultured onPLA/PGA scaffolds increased when the macroporesize was below 160 lm. However, the results of in vivostudies of fibrovascular tissue ingrowth using PVAsponges with pore size of 250 lm33 and poly(lactide-co-glycolide) tubular scaffolds (50–300 lm pore size)6

recommend increasing the pore size of PCL graftsto promote tissue invasion and integration with hosttissue.

The present study demonstrated that micro/macro-porous PCL tubes can accommodate the hydrody-namic stresses generated by short-term, simulatedblood flow conditions and the material shows goodpotential for integration with host fibrovascular tissue.

FIGURE 9. Proliferation rates of 3T3 fibroblasts on TCP and13% macroporous PCL graft material having specific pore sizeranges of 45–90, 90–125, and 125–250 lm (*Statistically sig-nificant difference between means at the 99% confidenceinterval p < 0.01, n 5 3).

FIGURE 10. Scanning electron micrographs of (a) surface ofmacroporous PCL graft material, (b) Swiss 3T3 mouse fibro-blasts cultured for 1 day, and (c) 3 days on 13% macroporousPCL material.


Further testing is required to quantify transmural flowrates in macroporous tubes and to establish whethersealed samples are able to withstand hemodynamicarterial pressures in the absence of transmural flow.Investigations of the long-term biomechanical prop-erties under blood flow conditions and the interactionof PCL tubes with blood components are necessary tofully assess whether micro/macroporous PCL tubescan provide a scaffold for regenerating tissue to build afunctional blood vessel as the PCL phase graduallyresorbs. A required patency of at least 7–10 yearsrecommends investigations of EC seeding of theluminal tube surface, for example, to counteractplatelet deposition and pseudointimal hyperplasia.

CONCLUSIONS

Microporous PCL tubes (nominal outer diameter7.5 mm, internal diameter 5.5 mm) designed for vas-cular tissue engineering can withstand the hydrody-namic stresses induced by short duration flow rates upto 1000 mL/min, which result in estimated internalpressures in excess of arterial pressure (80–130 mmHg).Micro/macroporous PCL tubes having a maximummacroporosity of 23% can accommodate the hydro-dynamic stresses generated by short duration, flowrates up to 1000 mL/min, which result in esti-mated internal pressures similar to venous pressure(30 mmHg). Both microporous and macroporous tubesexhibit burst strengths in excess of 900 mmHg which ishigher than arterial blood pressure by a factor ofapproximately 7. The materials are efficiently colonizedby fibroblasts in cell culture, indicating the potential forintegration with host fibrovascular tissue. These prop-erties recommend further investigations of cast, micro/macroporous PCL materials as scaffold structures forvascular tissue engineering.

REFERENCES

1Bittinger, F., S. Muhlbayer, V. Salih, M. Grau,H. A. Richter, C. L. Klein, et al. Experimental studies onthe suitability of human mesothelial cells for seeding vas-cular prostheses: shear stress resistance in vitro. J. Mater.Sci. Mater. Med. 7:543–550, 1996.2Chang, H. I., Y. Perrie, and A. G. Coombes. Delivery ofthe antibiotic gentamicin sulphate from precipitation castmatrices of polycaprolactone. J. Control Release110(2):414–421, 2006.3Cheng, C. P., D. Parker, and C. A. Taylor. Quantificationof wall shear stress in large blood vessels using Lagrangianinterpolation functions with cine phase-contrast magneticresonance imaging. Ann. Biomed. Eng. 30:1020–1032, 2002.

4Coombes, A. G., S. C. Rizzi, M. Williamson, J. E. Barralet,S. Downes, and W. A. Wallace. Precipitation casting ofpolycaprolactone for applications in tissue engineering anddrug delivery. Biomaterials 25(2):315–325, 2004.5Darling, A. L., and W. Sun. 3D microtomographic char-acterization of precision extruded poly-epsilon-capro-lactone scaffolds. J. Biomed. Mater. Res. B Appl. Biomater.70B(2):311–317, 2004.6Day, R. M., A. R. Boccaccini, V. Maquet, S. Shurey, A.Forbes, S. M. Gabe, et al. In vivo characterization of anovel bioresorbable poly(lactide co-glycolide) tubular foamscaffold for tissue engineering applications. J. Mater. Sci.Mater. Med. 15:729–734, 2004.7Gordon, J. E. Structures. Cambridge: Da Capo Press,2003.8Greisler, H. P., D. U. Kim, J. B. Proice, and A. B.Voorhee. Arterial regenerative activity after prostheticimplantation. Arch. Surg. 120:315–323, 1985.9Greissler, H. P., J. Ellinger, T. H. Schwarcz, J. Golan,R. M. Raymond, and D. U. Kim. Arterial regenerationover a polydioxanone prosthesis in the rabbit. Arch. Surg.122:715–721, 1987.

10Guidoin, R., D. Marceau, J. Couture, T. J. Rao, Y. Merhi,P.-E. Roy, and D. De la Feye. Collagen coatings as bio-logical sealants for textile arterial prostheses. Biomaterials10:156–165, 1989.

11Ku, B. K., and R. C. Allen. Vascular grafts. In:The Biomedical Engineering Handbook, edited by J. D.Bronzino. Boca Raton: CRC Press, 2006, pp. 128-121–128-128.

12L’Heureux, N., S. Paquet, R. Labbe, L. Germain, andF.A.Auger.Acompletelybiological tissue-engineeredhumanblood vessel. [see comment]. FASEB J. 12(1):47–56, 1998.

13Lee, S. J., J. Liu, S. H. Oh, S. Soker, A. Atala, andJ. J. Yoo. Development of a composite vascular scaffoldingsystem that withstands physiological vascular conditions.Biomaterials 29:2891–2898, 2008.

14Matsumura, G., N. Hibino, Y. Ikada, H. Kurosawa, andT. Shin’oka. Successful application of tissue engineeredvascular autografts: clinical experience. Biomaterials24(13):2303–2308, 2003.

15Mattern, K. J., C. Nakornchai, and W. M. Deen. Darcypermeability of agarose glycosaminoglycan gels analysedusing fiber mixture and donnan modes. Biophys. J. 95:648–656, 2008.

16Moore, G. F., and S. M. Saunders. Advances in biode-gradable polymers. Rapra Rev. Rep. 9:19–21, 1997.

17Mujais, S. K. Non thrombogenic surfaces. In: Blood-Material Interaction. A Basic Guide from Polymer Scienceto Clinical Application, edited by D. Falkenhagen,H. Klinkmann, E. Piskin, K. Opatmy. Glasgow: INFA,1998, 60–64.

18Niklason, L. E., J. Gao, W. M. Abbott, K. K. Hirschi, S.Houser, R. Marini, et al. Functional arteries grownin vitro. Science 284(5413):489–493, 1999.

19Papadaki, M., and L. V. McIntyre. Quantitative measure-ment of shear stress effects on endothelial cells. In: TissueEngineering Methods and Protocols, edited by J. R.Morgan, and M. L. Yarmsh. Totowa, NJ: Humana PressInc., 1999, pp. 577–592.

20Phaneuf, M. D., W. C. Quist, F. W. LoGerfo, andM. J. Bide. Surface modification of Dacron vascular grafts:incorporation of antithrombin and mitogenic properties.In: Polymeric Biomaterials, edited by S. Dumitriu. NewYork: Marcel Dekker, 2001, pp. 647–668.

WANG et al.214

21Schneck, D. J. An outline of cardiovascular structure andfunction. In: The Biomedical Engineering Handbook, edi-ted by J. D. Bronzino. Boca Raton: CRC Press, 2006,pp. 1.1–1.12.

22Shinoka, T., D. Shum-Tim, P. X. Ma, R. E. Tanel, N.Isogai, R. Langer, et al. Creation of viable pulmonaryartery autografts through tissue engineering. J. Thorac.Cardiovasc. Surg. 115:536–546, 1998.

23Tada, S., and J. M. Tarbell. Fenestral pore size inthe internal elastic lamina affects transmural flow distri-bution in the artery wall. Ann. Biomed. Eng. 29:456–466,2001.

24Van Damme, H., M. Deprez, E. Creemers, and R. Limet.Intrinsic structural failure of polyester Dacron vasculargrafts. A general review. Acta Chir. Belg. 105:249–255,2005.

25Vohra, R. K., G. J. Thomson, H. M. Carr, H. Sharma, M.Welch, and M. G. Walker. In vitro adherence and kineticstudies of adult human endothelial cell-seeded polytetra-fluoroethylene and gelatin-impregnated Dacron grafts. Eur.J. Vasc. Surg. 5:93–103, 1991.

26Walsh, P. W., S. Chin-Quee, and J. E. Moore. Flowchanges in the aorta associated with the deployment of anAAA stent graft. Med. Eng. Phys. 25:299–307, 2003.

27Wang, Y., H. I. Chang, D. Wertheim, A. C. Jones, C.Jackson, and A. Coombes. Characterisation of themacroporosity of polycaprolactone-based biocompositesand release kinetics for drug delivery. Biomaterials28:4619–4627, 2007.

28Wang, Y., P. Tomlins, A. Coombes, and M. Rides. On thedetermination of Darcy permeability coefficients for a

microporous tissue scaffold. Tissue Eng. C Methods2:281–289, 2010.

29Williams, J. M., A. Adewunmi, R. Schek, C. L. Flanagan,P. H. Krebsbach, S. E. Feiberg, et al. Bone tissue engi-neering using polycaprolactone scaffolds fabricated viaselective laser sintering. Biomaterials 26:4817–4827, 2005.

30Williamson, M. R., R. Black, and C. Kielty. PCL-PUcomposite vascular scaffold production for vascular tissueengineering: attachment, proliferation and bioactivity ofhuman vascular endothelial cells. Biomaterials27(19):3608–3616, 2006.

31Williamson, M. R., K. J. Woollard, H. R. Griffiths, andA. G. Coombes. Gravity spun polycaprolactone fibers forapplications in vascular tissue engineering: proliferationand function of human vascular endothelial cells. TissueEng. 12(1):45–51, 2006.

32Xue, L., and H. P. Greisler. In principles of tissue engineer-ing. In: Blood Vessels, edited by R. P. Lanza, and R. J. V.Langer. San Diego: Academic Press, 2000, pp. 427–446.

33Yamamoto, M., Y. Tabata, N. Kawasaki, and Y. Ikada.Promotion of fibrovascular tissue ingrowth into poroussponges by basic fibroblast growth factor. J. Mater. Sci.Mater. Med. 11:213–218, 2000.

34Yang, J., G. Shi, J. Bei, S. Wang, Y. Cao, Q. Shang, et al.Fabrication and surface modification of macroporouspoly(L-lactic acid) and poly(L-lactic co glycolic acid)(70/30) cell scaffolds for human skin fibroblasts cell culture.J. Biomed. Mater. Res. 62:438–446, 2002.

35Yao, H., and W. Y. Gu. Convection and diffusion incharged hydrated soft tissues: a mixture theory approach.Biomech. Model. Mechanobiol. 6:63–72, 2007.


Biomechanical Characterization of a Micro/Macroporous Polycaprolactone Tissue Integrating Vascular Graft

Documents