Biomaterials 24 (2003) 2549–2557

In vivo biostability of a poly(carbonate-urea)urethane graft

Alexander M. Seifaliana,*, Henryk J. Salacinskia, Alok Tiwaria, Alan Edwardsb,Staffan Bowaldc, George Hamiltona

aTissue Engineering Centre, University Department of Surgery, Royal Free and University College Medical School, University College London, Royal

Free Hospital, London NW3 2QG, UKbCredent Vascular Technologies Ltd., Cambrian House, Harwd Road, Brymbo, Wrexham LL5EA, UK

cUniversity Hospital, University of Uppsala, Uppsala, Sweden

Received 11 July 2002; accepted 11 December 2002

Abstract

In peripheral and coronary bypass surgery, the patency of prosthetic grafts is inferior to autologous vein, mainly due to intimal

hyperplasia caused in part by compliance mismatch between rigid graft and elastic host artery. We have developed a compliant

poly(carbonate-urea)urethane vascular graft ‘‘MyoLinkt’’ which was biostable in vitro degradation studies. To further investigate

the biostability of this material, we report a long-term in vivo study on 8 beagle dogs (1573 kg) implanted with this graft (ID 5mm)

in the aorta–iliac position; three grafts were harvested at 18 months to assess short-term biodegradation, with one animal having

died from an unrelated infection. The 4 remaining grafts were harvested at 36 months for analysis by: (1) histology, (2) compliance

measurements and (3) environmental scanning electron microscopy (ESEM); gel permeation chromatography (GPC); attenuated

total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) and radial tensile strength analysis.

There was no infection or inflammation of the grafts or surrounding tissues. Histological analysis showed a well-developed

neointima but only at the distal anastomosis. There were no significant differences in compliance pre- and post-implantation and no

evidence of material curvature, radial expansion or chemical breakdown, ESEM and GPC showed no signs of degradation. Peak

height analysis with ATR-FTIR of the 1740 cm�1 (C¼O of carbonate) and 1253 cm�1 bands (C–O–C of CO–O–C) showed a loss of

carbonate carbonyl but was not statistically significant. Radial tensile strength remained within batch release specifications.

This polyurethane graft retains its compliance post-implantation, whilst exhibiting only a minor hydrolysis of the amorphous

segment, confirming its biostability in vivo up to 3 years.

r 2003 Elsevier Science Ltd. All rights reserved.

1. Introduction

Autologous saphenous vein still remains the conduitof choice for below-knee vascular reconstructions.However in around 30% of such patients saphenousvein is unavailable and surgeons have had to rely eitherupon an arm vein or a polymeric non-biological graftmaterial principally expanded polytetrafluoroethylene(ePTFE) or polyethylene terephthalate (Dacron) [1,2].The patency rate of these prosthetic materials is inferiorin below knee anastomoses due to intimal hyperplasia(IH) particularly at the distal anastomosis [3]. It is thehypothesis of this group and others that IH is caused bythe compliance mismatch between inelastic graft and the

elastic artery at the anastomosis [4,5]. This resulting inincreased mechanical stimulus around the area of theanastomosis so generating turbulent and disturbed flow,this subsequently damaging the endothelial cell liningand causing the proliferation of smooth muscle cells andso the presence of IH [5].

There has been considerable interest and activity indeveloping alternative materials whose elasticity is closeto that of the host tissue to allow compliance matchingthis focussing especially on the chemical compositionsand different fabrication designs of polyurethanes [5,6].

Polyurethanes are generally copolymers consisting ofcrystalline (hard) and amorphous (soft) (macromono-mer) segments [6]. The record of these implants has sofar been mixed clinically because of the tendency tobiodegrade by hydrolytic and oxidative mechanisms[7,8]. Generally the site of the biodegradation is in thesoft segment typically composed of ester or ether

*Corresponding author. Tel.: +44-20-7830-2901; fax: +44-20-7472-

6444.

E-mail address: a.seifalian@rfc.ucl.ac.uk (A.M. Seifalian).

0142-9612/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved.

doi:10.1016/S0142-9612(02)00608-7

compounds. Whilst hydrolysis of the urethane (or urea)hard segment linkages is possible, those groups arerelatively stable and do not comprise the primary modeof biodegradation. Grafts made from poly(ester) softsegments have generally degraded quickly and severelyby hydrolysis [9] whilst poly(ether) segments have beensusceptible to oxidative degradation, but this beingproportional to ether content [10–15]. In associationwith an industrial concern (Credent Vascular Technol-ogies Ltd.), we have developed a polyurethane graftbased on poly(carbonate) soft segments trade marked‘‘MyoLink’’.

Obtaining long-term compliance post-implantationhas been difficult as to date polyurethane grafts haverelied on overall dilation that is rapidly negated post-implantation by peri-vascular ingrowth. The designapproach used here has been to develop a prosthesisthat maintains compliance and pulsatile flow in vivo byenabling the transmission of energy and a better qualityof flow. This has been achieved by the incorporation of ahoneycomb structure by usage of a cast coagulationextrusion at room temperature [16].

During phased coagulation the fillers, Tween 80 andsilicon dioxide prevent collapse of the structure as thesolvent disperses and the filler dissolves into thecoagulant, resulting in a single layer uniform micro-porous structure. The incorporation of this honeycombstructure allows luminal expansion through a mechan-ism of internal wall compression without the need forexternal dilation. This methodology allows compliancematching of the graft to the host artery [17,18], and for asuperior transmission of pulsatile blood flow.

MyoLink has undergone in vitro degradation studiesinitially using a glass wool, hydrogen peroxide andcobalt chloride system where the grafts were pre-stressedover an ePTFE mandril to achieve 50% elongation [16]and then hydrolytic, peroxidative and blood basedbiological systems [19,20] with no pre-stressing in whichonly minor traces of degradation where evident. Toconfirm these studies and assess the biostability ofMyoLink in terms of its mechanical properties andmaterials chemistry a long-term (36 months) canineimplant study was performed.

2. Methods and materials

2.1. Graft materials

Details of the synthesis of the graft materialMyoLinkt have been published previously[5,16,19,21,22], therefore only a brief summary will beprovided here. The chemistry is based on methylenediisocyanate, polycarbonate diol and ethylene diamineall dissolved in dimethylacetamide in molar ratios of2:1:0.97 [21,16]. As this polymer solution is extruded

onto a rotating mandril head into coagulant themicroporous structure is generated.

2.2. Surgical model-implant technique

Eight adult beagle dogs’ weighed at 1573 kg wereanaesthetised by IV injection of sodium pentothal 8mg/kg body weight. After tracheal intubation the animalswere connected to a Siemens 900C ventilator andanesthesia was continued with a mixture of oxygen/nitrous oxide/halothane gasses in the ratio of 40/59/1%at a ventilation rate of 4–6 l/min.

The abdomen was shaved and swabbed with hibitanein alcohol before the application of sterile drapes. Alaparotomy was performed and the aorta, iliac and renalarteries were exposed, 1000 units of heparin wereadministered intravenously before partial longitudinalclamping of the infra renal aorta. A longitudinal incisionapproximately 2–2.5 times the diameter of the graft wasmade and a single MyoLink graft of 5-mm internaldiameter (ID) was cut to fit the aortic incision typically60mm in length. An end-to-side anastomosis wascreated using double-ended 5/0 Prolene sutures. Staysutures were placed at the toe and heel of theanastomosis to prevent ‘‘a purse string effect’’ and eachside was completed with continuous sutures. The graftwas cut to length to accommodate an approximate 10%stretch and the procedure was repeated to create end-to-side anastomoses to the iliac arteries with 5/0 Prolenesutures.

The aorta was ligated proximal to the distal anasto-moses and the abdomen was closed with polydioxononeloop and sutures (Ethicon). Post-operatively the animalswere given procainpenicillin 5 cc daily for one week. Allanimals survived surgery. Three animals were electivelysacrificed after 18 months to assess short-term biode-gradation. One animal died from an unrelated infection.The remaining four animals were sacrificed after 36months and the grafts removed. They were divided intothree groups of segments for: (1) histological study, (2)compliance measurements in which, the outer perivas-cular tissue was still present, and (3) materials chemistryand mechanical measurements for which the outerperivascular tissue was carefully removed these beingas follows: Dimensional and visual characterisation;Environmental scanning electron microscopy (ESEM);Gel permeation chromatography (GPC); AttenuatedTotal Reflectance-Fourier Transform infrared spectro-scopy (ATR-FTIR) and radial tensile strength analysis.In the latter procedure the grafts were washed withsterile water and the surrounding tissue capsule wasremoved by blunt dissection. Then the capsule wasseparated from the outer surface of the graft. Therewas firmer attachment of the perivascular tissue at theexternal suture lines and the tissue was left undisturbedto avoid damage to the graft surface. The grafts were

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–25572550

then placed in a 5% w/w solution of tergazyme tochemically digest the tissue that had grown into thehoneycomb structure of the graft. The solution waschanged every 72 h and after 16 days the explants werecompletely clean of surrounding tissue. Post-chemicaltreatment the grafts were washed in sterile water anddried in an oven at 40�C for 24 h.

2.3. Histology of explanted samples

An independent investigator performed all of thehistological analyses. The graft segments were presentedin 2% formalin solution. Standard staining pro-tocols were used: hematoxylin-eosin (H&E); Trichrome(TRI) and Phosphotungstic acid hematoxylin (PTAH)stains.

2.4. Compliance measurements

The explanted grafts for compliance studies wereremoved with surrounding tissue and placed in salinesolution. In vitro dynamic compliance measurements ata range of constant pulse pressure of 60mmHg over arange of mean pressure from 30 to 100mmHg wasundertaken. To simulate the physiological flow, a flowcircuit specifically designed accurately to reproduce thehaemodynamic characteristics of pulsatile arterial bloodflow was fabricated. This was then used to perfusesegments of the vascular grafts with blood for com-pliance measurements. The technique has been describedpreviously [5,18,23,24]. Briefly the flow model compriseda variable-speed electromagnetic centrifugal pump (BioMedicus, Minnetonka, MN, USA), flexible plastictubing and a fluid reservoir. A previously developedgovernable flow waveform conditioner was sited inseries with the circuit and used to generate arterial flowwaveforms. Instantaneous flow rate was measured usinga 6-mm tubular flow probe connected to a TransonicMedical flowmeter system (HT207; Transonic MedicalSystem, Ithaca, NY, USA). Serial intraluminal pressuremeasurements were made at discrete sites along the graftusing a Millar Mikro-tip catheter transducer (MillarInstruments, Houston, Texas, USA) introduced via aY-connection port. The perfusion solution used washeparinised human blood from the hospital blood bank.A gas mixture of 95% oxygen and 5% carbon dioxidewas bubbled through the blood reservoir and theperfusate was maintained at 37�C.

Grafts were mounted in series in the flow circuit,exposed to flow and placed in a bath containing normalsaline maintained at 37�C. Segments were subject to10% longitudinal stretch in order to reproduce thelikely in vivo conformation. All outputs were fed into acommercial analogue-to-digital data acquisition record-ing system (ADC/MacLab; AD Instruments, Hastings,UK).

The change in vessel wall diameter with respect toeach cardiac cycle was measured at discrete sites alongthe graft, with measurements taken in the sagittal planeat 90� to the long axis of the graft. Segments of the graftwere scanned with a specially adapted, duplex ultra-sonography system (ATL 5000), with signal output to ahigh-resolution, echo-locked wall-tracking system (WallTrack; Pie Medical Systems, Netherlands). This systemallowed the measurement of vessel wall movement overtime by automatically tracking assigned points of theinduced radiofrequency signal deemed to be representa-tive of the anterior and posterior arterial wall. Adetailed description of the system has been previouslypublished [18]. Measurements were made using aconstant pulse pressure over a range of mean pressures.For each mean pressure and test segment, threemeasurements of wall movement were made.

Diametrical compliance (C), defined as the inverse ofPetersen’s elastic modulus Ep and expressed as%mmHg�1� 10�2, was determined for each test seg-ment at each mean pressure as follows: C ¼ ½ðDs �DdÞ=Dd� � 104=ðPs � PdÞ where Ds and Dd are systolicand diastolic diameter, and Ps and Pd are systolic anddiastolic pressure, respectively.

2.5. Materials assessment of vascular grafts post-

biodegradation

2.5.1. Dimensional and visual characterisation

Effects of biodegradation on dimensional character-istics: ID; wall thickness and visual examination wereperformed to International Standards Organisation(ISO) standard operating procedures and certificationstandards. The ID of the grafts was measured using acalibrated taper gauge (8.5-ISO/DS7198/40). Measure-ments of graft wall thickness were performed with aconstant load gauge (8.7-ISO/DS7198) and a visualassessment was made with white light back illumination(8.1-ISO/DS7198).

2.5.2. Environmental scanning electron microscopy

The MyoLink grafts pre- and post-explanted wereplaced inside the ESEM instrument. The ESEMmicrographs were taken at 1.5mm intervals along thegrafts. The digital micrographs were ‘blind’ examinedfor features indicating biodegradation such as fissuring,interlocking, cracking and surface grazing of the graftmaterial by a professional independent assessor.

2.5.3. Gel permeation chromatography

An independent assessor also carried out GPCanalysis. The average molecular weight and the poly-dispersity of the graft samples were measured using aWaters GPC system (Waters Ltd., Ont, Canada)equipped with a refractive index detector and dataacquisition system. Dimethylformamide (DMF, Aldrich

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–2557 2551

Ltd., UK) with 0.05m of lithium bromide (LiBr, AldrichLtd., UK) was used as the mobile phase in conjunctionwith a set of four columns (Styragel, HT1-4, USA). Thetemperature of the columns was 80�C, and the flow ratewas 1ml/min. The molecular weight values werereported as polystyrene equivalent molecular weightvalues.

2.5.4. Attenuated total reflectance-Fourier transform

infrared spectroscopy (ATR-FTIR)

ATR-FTIR spectra of the treated and untreatedmaterials were obtained with a Nicolet 800 FTIRspectrometer using a germanium crystal with a 60�Cend-face angle equivalent to the incidence IR beam.Each spectrum was normalised to the 1591 cm�1 bandaromatic n(C¼C) as this band is considered an internalreference peak because it has been shown to beunaffected during short term in vivo implantationpreviously [16]. Peak height analysis was done on the1740 cm�1 n(C¼O) band associated with the carbonatelinkage [25]. The band at 1253 cm�1 was assigned ton(C–O–C) of (CO–O–C) and also was examined forchanges due to hydrolysis of the carbonate linkage.Data were recorded pre-implant and at post-implant atthe following points along the graft: distal anastomosesand at 1, 2 and 3mm away from the anastomosis.

2.5.5. Radial tensile strength

Radial tensile strength testing was performed tostandard operating procedures on a tensilometer (In-stron 1011, Instron Ltd., High Wycombe, Bucks, UK).The test piece was 1270.5mm and was threaded over asplit pin assembly and loaded into the Instron machinewhere the separation of the jaw was performed at rate of50mm/min using the American Society for Testing andMaterials (ASTM 1708) method and ISO 527. Theextension and peak load was then recorded and load atbreak calculated from the following equation:

Load per unit length

¼ Maximum load=ð2� length of sampleÞ:

This resulted in stress–strain curves, which are given inunits of N/mm.

3. Results

Total autopsies performed on all of the animalsshowed they were all healthy apart from the one animalwhich died of an unrelated infection. No signs of grossinfection and in particular inflammation of the grafts orsurrounding areas were evident. Macroscopically therewere no indications of material curvature, radialexpansion or chemical breakdown. Furthermore in noneof the animals was there any evidence of aneurysmformation.

3.1. Histological results

Fig. 1 shows a typical post-implantation MyoLink graftexplanted after 36 months. The H&E; TRI and PTAHstains (� 6.6 and � 13.2) respectively, all showed a well-developed neointima lining the distal anastomotic regionof the graft arising from the adjacent artery, which extendsinto the graft but not the main body of the graft itself(Figs. 2 and 3). Fig. 2 shows representative H&E, TRI andPTAH (� 6.6) stains of the graft sectioned in the coronalplane in the distal anastomotic zone. In this region therewas B0.55mm of neointima (graft wall thickness being0.9mm) this translating to 36% decrease in cross-sectionalarea. At 1–3mm distances away from the anastomosisthere was a decrease in neointima thickness with nonebeing present in the main body of the graft.

Furthermore extensive collagen deposition (in theform of collagen bundles) and cellular infiltration intothe honeycomb structure of the graft was clearly evidentwith the surface of the neointima being covered by alayer of endothelial cells (Fig. 3). However, there was noperipheral haematoma around any of the grafts.

Fig. 1. (a) Graft post-implantation, and (b) with suture demonstrating

the tensile strength of MyoLink graft post-implant.

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–25572552

3.2. Compliance

The results of compliance measurements made on thegrafts as a function of pressure at a selected position andalong the graft’s length at constant physiologicalpressure are depicted in Fig. 6. Overall, there were nosignificant differences in the compliance values obtainedfor MyoLink grafts for measurements made at bothmean pressures pre- and post-implantation (P > 0:1).For comparison the compliance of ePTFE and Dacronhave been included over this pressure range andit is immediately evident that the difference in viscoe-lasticity is large even in the MyoLink graft post-implantation.

3.3. Dimensional and visual characterisation

This showed that the dimensions and appearance ofthe grafts remained within normal batch release limits(Table 1).

Fig. 2. Light micrographs (6.6� ) of: (a) H&E, (b) TRI, and (c) PTAH

stains of MyoLinkt graft sectioned in the coronal plane in the

anastomotic zone explanted after 36 months.

Fig. 3. Light micrographs (13.2� ) of: (a) H&E, and (b) TRI of

MyoLinkt graft sectioned in the coronal plane in the anastomotic

zone explanted after 36 months. These show architectural remodelling

had occurred as evidenced by extensive collagen deposition and

cellular infiltration into the honeycomb structure of the graft.

Table 1

Summary of results of dimensional and visual characterisation (DVC);

radial tensile strength; gel permeation chromatography (GPC) and

attenuated total reflectance Fourier transform infrared spectroscopy

(ATR-FTIR)

Pre-implant Post-implant P

DVC

Internal diameter (mm) 570.04 570.11 NS

Wall thickness (mm) 0.970.03 0.970.05 NS

Radial tensile strength (N/mm) >1 1.4870.07 NS

GPC

MN 4530072900 4160074500 NS

MW 9850073500 10400077500 NS

D 2.1770.32 2.4970.33 NS

ATR–FTIR 1740 band (cm�1)

At anastomosis 1.4570.07 1.1770.09 NS

1mm away 1.4570.07 1.2070.08 NS

2mm away 1.4570.07 1.3370.05 NS

3mm away 1.4570.07 1.3570.04 NS

NS—not significant.

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–2557 2553

3.4. Environmental scanning electron microscopy

ESEM showed no signs of pitting and etching knownindicators of degradation (see Fig. 4), and so no loss ofbiostability.

3.5. Gel permeation chromatography

GPC showed there was no statistical differencebetween control and implanted samples typical valuesbeing given in Table 1.

3.6. Attenuated total reflectance-Fourier transform

infrared spectroscopy

The poly(carbonate-urea)urethane in this study has apoly(carbonate) amorphous segment where the majorabsorption bands were present at the two followingwavelengths 1740 cm�1 (C¼O of the carbonate)and 1253 cm�1 (C–O–C this being from CO–O–C). Inthis particular polymer the 1740 cm�1 band is consid-ered to be primarily due to the carbonate carbonylalthough the urethane carbonyl contributes slightly [25].

Fig. 4. Typical ESEM pictures of MyoLinkt graft post-implant, (a) at the anastomoses and (b) inner surface. The pores visible on the surface of

graft are not environmental stress cracking (ESC) or a form of degradation, but a natural feature induced during the manufacturing process to form a

compliant graft (figures are originally magnification � 10 and 20).

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–25572554

ATR-FTIR peak height analysis of the 1740 cm�1

(C¼O of carbonate) showed a loss of carbonatecarbonyl post-implantation. Similar trends were seenin the 1253 cm�1 (C–O–C of CO–O–C) bands, but inboth cases this was not statistically significant. Fig. 5shows a typical ATR-FTIR spectrum from a post-implant MyoLink graft with the decreases in peakheight being attributed to a gradual hydrolysis of thecarbonate linkage.

3.7. Radial tensile strength analysis

Tensile testing showed all samples remained withinthe batch release specifications, which is greater than1N/mm (1.4870.07).

4. Discussion

It is logical that any vascular prosthesis should befabricated to mimic as far as possible the mechanicalcharacteristics of natural artery. Obtaining long-termcompliance has been an elusive goal, because grafts todate have relied upon overall external dilation which israpidly compressed by perivascular in growth, resultingin constriction and loss of compliance [26].

Polyurethanes have been investigated as an alterna-tive polymer for the fabrication of small diametervascular prostheses (o5mm), due to their elasticproperties and ability to be generated by a variety ofprocessing methods [6]. Two principal difficulties havelimited the clinical application of virtually all polyur-

ethane vascular grafts: poor biostability and loss ofcompliance post-implantation.

A polyurethane is the reaction product of anisocyanate (or diisocyanate), macroglycol and chainextender [6]. The isocyanate is reacted with the macro-glycol to form a low molecular weight pre-polymer.Then a chain extender is added to produce a highmolecular weight polyurethane. It is the macroglycol(soft segment) that imparts the elasticity to the urethane.Our group has developed a polyurethane in which thesoft segment is based on a carbonate group which ishighly resistant to oxidation. This resistance has beenconfirmed by in vitro studies [19] with and without pre-stressing. However, these studies suggested gradualhydrolysis of the carbonate segment occurs under someconditions [20].

To evaluate the materials biostability and assesswhether compliance is maintained in vivo post-implan-tation a long-term (36 month) animal study wasimplemented. Surface analysis of the explanted graftby ATR-FTIR showed a small decrease in carbonateamorphous segment peak height implying some micro-phase separation of the soft and hard segment hadoccurred, but importantly no frequency shift of theamine peak was observed showing the urethane linkageswere completely intact. ESEM, and tensile strengthanalysis measurements showed that physical changesthat typically occur on implantation with ester and etherbased polyurethanes (swelling, plasticisation, crystal-lisation/decrystallisation, fatigue fracture, creep, stresscracking and kinking) were not present in this poly-urethane graft. These findings support previous in vitro

Fig. 5. A typical ATR/FTIR spectra of MyoLink graft, pre-implantation (control) at distal anastomoses and at 1, 2 and 3mm away from

anastomoses post-implantation. There were no statistically significant changes in spectrums peaks at wavelength 1740, and 1253 cm�1 of the

carbonate soft segments.

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–2557 2555

degradation tests and those of a short-term animal study[16,20]. However importantly this study shows that thecompliance of the graft does not change following theperivascular reaction that is commonly seen withconventional prosthetic grafts. In one study of com-pliance with ePTFE based vascular grafts this wasreduced to 14% and with Dacron to 29% of their pre-implantation levels after 3 months of implantation [27](Fig. 6). Although in this study, there was in the order ofa 6% reduction in compliance for the MyoLink graftafter 36 months, this value was not statisticallysignificant.

The other main advantage of this graft is its excellentsurface characteristic particularly when used for tissueengineering where it shows significantly better resultsthan ePTFE grafts [28,29]. Our goal in tissue engineer-ing has been to reduce the thrombogenicity of thegraft [30].

In vivo the honeycomb structure of Myolink in thisstudy has been shown to facilitate the development of anendothelialised internal capsule coupled with infiltrationof tissue into the graft wall. This latter effect isparticularly important because it prevents pannus and/or tissue build-up at the anastomotic sites as well as itspossible detachment from the surface, two frequentlyencountered causes of graft thrombosis in non-porouspolyurethane based vascular grafts [31,26].

5. Conclusion

This study has demonstrated that the complianceproperties of this vascular graft are retained post-

implantation and that only a minor loss of theamorphous segment occurs (this not being statisticallysignificant) so it can be concluded that a bio-stable polyurethane graft material has been developed.Currently this graft is undergoing a phase I clinicaltrial.

Acknowledgements

The authors acknowledge Professor James M. An-derson, The Institute of pathology of Case WesternReserve University, USA for the histology report, Dr.Marianne Odlyha, Department of Chemistry, BirkbeckCollege, University of London, London, UK for theATR-FTIR analysis.

We would like to acknowledge the financial supportfrom Credent Vascular Technology Ltd., UK whoprovided a grant to support HJS and AT and thePublic Health grant M142 & M174 which providedfunds for laboratory equipment.

References

[1] Seifalian AM, Tiwari A, Hamilton G, Salacinski HJ. Improving

the clinical patency of prosthetic vascular and coronary bypass

grafts: the role of seeding and tissue engineering. Artif Organs

2002;26:307–20.

[2] Faries PL, LoGerfo FW, Arora S, Hook S, Pulling MC, Akbari

CM, Campbell DR, Pomposelli Jr FB. A comparative study of

alternative conduits for lower extremity revascularization: all-

autogenous conduit versus prosthetic grafts. J Vasc Surg 2000;

32:1080–90.

[3] Wu MH, Shi Q, Sauvage LR, Kaplan S, Hayashida N, Patel MD,

Wechezak AR, Walker MW. The direct effect of graft compliance

mismatch per se on development of host arterial intimal

hyperplasia at the anastomotic interface. Ann Vasc Surg 1993;7:

156–68.

[4] Abbott WM, Megerman J, Hasson JE, L’Italien G, Warnock DF.

Effect of compliance mismatch on vascular graft patency. J Vasc

Surg 1987;5:376–82.

[5] Salacinski HJ, Goldner S, Giudiceandrea A, Carson RJ, Edwards

A, Hamilton G, Seifalian AM. The mechanical behavior of

vascular grafts: a review. J Biomater Appl 2001;15:241–78.

[6] Lelah MD, Cooper SL. Polyurethanes in medicine. Florida: CRC

Press, 1986.

[7] Vermette P, Wang GB, Santerre JP, Thibault J, Laroche G.

Commercial polyurethanes: the potential influence of auxiliary

chemicals on the biodegradation process. J Biomater Sci Polym

Ed 1999;10:729–49.

[8] Griesser HJ. Degradation of polyurethanes in biomedical

applications. A review. Polym Degrad Stabil 1991;33:329–54.

[9] Pinchuk L. Polyurethane degradation: effects on mechanical

properties and carcinogenicity. In: Klitzman B, editor. Problems

in general surgery: biomaterial. Philadelphia: J.B. Lipincott, 1994.

p. 179–91.

[10] Szycher M. Surface fissuring of polyurethanes following in vivo

exposure. In: Fraker, Griffin, editors. ASTM STP 859. Philadel-

phia. 1983. p. 308–21.

[11] Stokes K, McVenes R, Anderson JM. Polyurethane elastomer

biostability. J Biomater Appl 1995;9:321–54.

Fig. 6. Compliance-mean pressure curves for pre- and post-implanta-

tion MyoLink grafts. ePTFE and Dacron data were plotted for Ref.

[18].

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–25572556

[12] Phua SK, Castillo E, Anderson JM, Hiltner A. Biodegradation

of a polyurethane in vitro. J Biomed Mater Res 1987;21:

231–46.

[13] Anderson J. Mechanisms of inflammation and infection with

implanted devices. Cardiovasc Pathol 1993;2:335–415.

[14] Zhao QH, McNally AK, Rubin KR, Renier M, Wu Y, Rose-

Caprara V, Anderson JM, Hiltner A, Urbanski P, Stokes K.

Human plasma alpha 2-macroglobulin promotes in vitro oxida-

tive stress cracking of Pellethane 2363-80A: in vivo and in vitro

correlations. J Biomed Mater Res 1993;27:379–88.

[15] Smith R, Williams DF, Oliver C. The biodegradation

of poly(ether urethanes). J Biomed Mater Res 1987;21:

1149–65.

[16] Edwards A, Carson RJ, Szycher M, Bowald S. In vitro and in

vivo biodurability of a compliant microporous vascular graft.

J Biomater Appl 1998;13:23–45.

[17] Tai NR, Giudiceandrea A, Salacinski HJ, Seifalian AM,

Hamilton G. In vivo femoropopliteal arterial wall compliance in

subjects with and without lower limb vascular disease. J Vasc Surg

1999;30:936–45.

[18] Tai NR, Salacinski H, Edwards A, Hamilton G, Seifalian AM.

Compliance properties of conduits used in vascular reconstruc-

tion. Br J Surg 2000;87:1480–8.

[19] Salacinski HJ, Tai NR, Carson RJ, Edwards A, Hamilton G,

Seifalian AM. In vitro stability of a novel compliant

poly(carbonate-urea)urethane to oxidative and hydrolytic stress.

J Biomed Mater Res 2002;59:207–18.

[20] Salacinski HJ, Odlyha M, Hamilton G, Seifalian AM. Thermo-

mechanical analysis of a compliant poly(carbonate-urea)urethane

after exposure to hydrolytic, oxidative, peroxidative and biologi-

cal solutions. Biomaterials 2002;23:2231–40.

[21] Salacinski HJ, Punshon G, Krijgsman B, Hamilton G, Seifalian

AM. A hybrid compliant vascular graft seeded with microvascular

endothelial cells extracted from human omentum. Artif Organs

2001;25:974–82.

[22] Tiwari A, Salacinski HJ, Seifalian AM, Hamilton G. New

prostheses for use in bypass grafts with special emphasis on

polyurethanes. Cardiovasc Surg 2002;10:191–7.

[23] Giudiceandrea A, Salacinski H, Tai N, Punshon G, Hamilton G,

Seifalian AM. Development and evaluation of an ideal flow

circuit: assessing the dynamic behavior of endothelial cell seeded

grafts. J Artif Organs 2000;3:16–24.

[24] Baguneid MS, Goldner S, Fulford PE, Hamilton G, Walker MG,

Seifalian AM. A comparison of para-anastomotic compliance

profiles after vascular anastomosis: nonpenetrating clips versus

standard sutures. J Vasc Surg 2001;33:812–20.

[25] Dillon JG. Infrared spectroscopic atlas of polyurethanes. Penn-

sylvania: Technomic Publishing Lancaster, 1989. p. 60–210.

[26] Marois Y, Akoum A, King M, Guidoin R, von Maltzahn W,

Kowligi R, Eberhart RC, Teijeira FJ, Verreault J. A novel

microporous polyurethane vascular graft: in vivo evaluation of

the UTA prosthesis implanted as infra-renal aortic substitute in

dogs. J Invest Surg 1993;6:273–88.

[27] Kidson IG, Abbott W. Low compliance and arterial graft

occlusion. Circulation 1978;58:I1–4.

[28] Tiwari A, Salacinski HJ, Hamilton G, Seifalian AM. Tissue

engineering of vascular bypass grafts: role of endothelial cell

extraction. Eur J Vasc Endovasc Surg 2001;21:193–201.

[29] Salacinski HJ, Tiwari A, Hamilton G, Seifalian AM. Cellular

engineering of vascular bypass grafts: role of chemical coatings

for enhancing endothelial cell attachment. Med Biol Eng Comput

2001;39:609–18.

[30] Tiwari A, Salacinski HJ, Punshon G, Hamilton G, Seifalian AM.

Development of a hybrid cardiovascular graft using a tissue

engineering approach. FASEB J 2002;16:791–6.

[31] Martz H, Paynter R, Ben Slimane S, Beaudoin G, Guidoin R,

Borzone J, Ben Simhon H, Satin R, Sheiner N. Hydrophilic

microporous polyurethane versus expanded PTFE grafts as

substitutes in the carotid arteries of dogs. A limited study.

J Biomed Mater Res 1988;22:63–9.

A.M. Seifalian et al. / Biomaterials 24 (2003) 2549–2557 2557