Biocompatibility and hemocompatibility of polyvinylalcohol hydrogel used for vascular grafting—In vitroand in vivo studies

Nuno Alexandre,1,2 Jorge Ribeiro,3,4 Andrea G€artner,3,4 Tiago Pereira,3,4

Irina Amorim,5,6 Jo~ao Fragoso,1 Ascens~ao Lopes,7 Jo~ao Fernandes,8,9

El�ısio Costa,8,10 Alice Santos-Silva,8,10 Miguel Rodrigues,7 Jos�e Domingos Santos,7

Ana Colette Maur�ıcio,3,4 Ana L�ucia Lu�ıs3,4

1Departamento de Zootecnia, Universidade de �Evora (UE), P�olo da Mitra, Apartado 94, 7002-554�Evora, Portugal2Instituto de Ciencias Agr�arias e Ambientais Mediterranicas (ICAAM), Universidade de �Evora (UE),

P�olo da Mitra, Apartado 94, 7002-554 �Evora, Portugal3Centro de Estudos de Ciencia Animal (CECA), Instituto de Ciencias e Tecnologias Agr�arias e Agro-Alimentares (ICETA),

Rua D. Manuel II, Apartado 55142, 4051-401 Porto, Portugal4Departamento de Cl�ınicas Veterin�arias, Instituto de Ciencias Biom�edicas de Abel Salazar (ICBAS),

Universidade do Porto (UP), Rua de Jorge Viterbo Ferreira, n� 228, 4050-313 Porto, Portugal5Departamento de Patologia e de Imunologia Molecular, Instituto de Ciencias Biom�edicas de Abel Salazar (ICBAS),

Universidade do Porto (UP), Rua de Jorge Viterbo Ferreira, n� 228, 4050-313 Porto, Portugal6Instituto Portugues de Patologia e Imunologia Molecular da Universidade do Porto (IPATIMUP),

Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal7CEMUC, Departamento de Engenharia Metal�urgica e Materiais, Faculdade de Engenharia da,

Universidade do Porto, s/n, 4200-465 Porto, Portugal8Instituto de Biologia Molecular e Celular (IBMC), Universidade do Porto (UP), Rua do Campo Alegre,

n� 823, 4150-180 Porto, Portugal9Instituto de Imagem Biom�edica e Ciencias da Vida (IBILI), Universidade de Coimbra (UC),

Azinhaga Santa Comba, Celas, 3000-548 Coimbra, Portugal10Laborat�orio de Bioqu�ımica, Departamento de Ciencias Biol�ogicas, Faculdade de Farm�acia,

Universidade do Porto (UP), Rua de Jorge Viterbo Ferreira, n� 228, 4050-313 Porto, Portugal

Received 15 January 2014; accepted 21 January 2014

Published online 14 February 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.35098

Abstract: Polyvinyl alcohol hydrogel (PVA) is a synthetic poly-

mer with an increasing application in the biomedical field that

can potentially be used for vascular grafting. However, the tis-

sue and blood–material interactions of such gels and mem-

branes are unknown in detail. The objectives of this study

were to: (a) assess the biocompatibility and (b) hemocompati-

bility of PVA-based membranes in order to get some insight

into its potential use as a vascular graft. PVA was evaluated

isolated or in copolymerization with dextran (DX), a biopoly-

mer with known effects in blood coagulation homeostasis.

The effects of the mesenchymal stem cells (MSCs) isolated

from the umbilical cord Wharton’s jelly in the improvement of

PVA biocompatibility and in the vascular regeneration were

also assessed. The biocompatibility of PVA was evaluated by

the implantation of membranes in subcutaneous tissue using

an animal model (sheep). Histological samples were assessed

and the biological response parameters such as polymorpho-

nuclear neutrophilic leucocytes and macrophage scoring

evaluated in the implant/tissue interface by International

Standards Office (ISO) Standard 10993-6 (annex E). According

to the scoring system based on those parameters, a total

value was obtained for each animal and for each experimental

group. The in vitro hemocompatibility studies included the

classic hemolysis assay and both human and sheep bloods

were used. Relatively to biocompatibility results, PVA was

slightly irritant to the surrounding tissues; PVA-DX or PVA

plus MSCs groups presented the lowest score according to

ISO Standard 10993-6. Also, PVA was considered a

Correspondence to: N. Alexandre; e-mail: nmla@uevora.ptContract grant sponsor: QREN I&DT Cluster in Development of Products for Regenerative Medicine and Cell Therapies (Projects Biomat & Cell

QREN 2008); contract grant number: 1372

Contract grant sponsor: European Community FEDER fund through North Portugal Regional Operational Program 2007–2013 (ON2—O Novo

Norte)

Contract grant sponsor: Fundac~ao para a Ciencia e Tecnologia (FCT), Minist�erio da Educac~ao e da Ciencia and Program Project Euronanomed,

Ref: EraNet—EuroNanoMed JTC2010; contract grant number: ENMED/0002/2010

Contract grant sponsor: Programa Operacional Factores de Competitividade (Program COMPETE); contract grant number: Project Pest-OE/AGR/

UI0211/2011

Contract grant sponsor: Nuno Alexandre has a Doctoral Grant from Fundac~ao para a Ciencia e Tecnologia (FCT), Minist�erio da Educac~ao e da

Ciencia, Portugal; contract grant number: SFRH/BD/64838/2009

4262 VC 2014 WILEY PERIODICALS, INC.

nonhemolytic biomaterial, presenting the lowest values for

hemolysis when associated to DX. VC 2014 Wiley Periodicals,

Inc. J Biomed Mater Res Part A: 102A: 4262–4275, 2014.

Key Words: mesenchymal stem cells, vascular graft, polyvinyl

alcohol hydrogel, hemocompatibility, biocompatibility, sheep

model

How to cite this article: Alexandre N, Ribeiro J, G€artner A, Pereira T, Amorim I, Fragoso J, Lopes A, Fernandes J, Costa E,Santos-Silva A, Rodrigues M, Santos JD, Maur�ıcio AC, Lu�ıs AL. 2014. Biocompatibility and hemocompatibility of polyvinylalcohol hydrogel used for vascular grafting—In vitro and in vivo studies. J Biomed Mater Res Part A 2014:102A:4262–4275.

INTRODUCTION

According to studies reported by the World Health Organiza-tion, one of the leading causes of death are diseases causedby arterial obstruction.1 The development of synthetic graftsfor vessel replacement has been extensively studied in thelast decades both as concerning the design of existing materi-als or the development of new synthetic materials.2 Biomedi-cal devices are required to be studied in terms ofbiocompatibility and hemocompatibility before pre-clinicaland clinical application as vascular grafts, according to stand-ards from International Standards Office (ISO) and Federaland Drug Administration. It is known that the surgicalimplantation of these devices results always in tissue injurythat can lead to inflammation and/or biointegration. Theassessment of hemocompatibility is a key aspect for medicaldevices such as artificial vascular grafts that contact directlyand continuously with blood. The degree of thrombogenicityof biomaterials used in the production of vascular graftsinfluence directly their function as a blood conduit due tothe formation of thrombus and consequently the magnitudeof their patency rate. According to the ISO standard 10993-4,hemocompatibility is also evaluated by in vitro tests thatpoint in evidence the blood–biomaterial interaction beforetheir use in animal models. Hemocompatibility describes theresponse of a material when contacts with the blood of thehost. A compatible response of the host can be different forthe same material in different applications; it can includeclotting, although this extreme response is not normally con-sidered blood-compatible. However, the healing of a vasculargraft is vastly facilitated, if a fibrin coating forms on thelumen of the graft at the anastomosis.3 Poor hemocompatibil-ity is related to a biomaterial-induced immunological andinflammatory responses including activation of coagulationand complement systems. As a direct result of the activationof the former systems, several side effects were observed,including anaphylactic reactions, increased risk of clot forma-tions, acute or chronic inflammation, infection, tumorigenesis,material degradation, renal and pulmonary dysfunction aswell as encapsulation and loss of function of the device. Theassessment of blood compatibility has several degrees ofimportance but it is vital in devices that contact directly withblood for different periods of time, including short-term (ascatheters), long-term (as heart valves and prosthetic vasculargrafts) contact and in extracorporeal devices (as dialysis andcardiopulmonary bypass machine).

Almost immediately after implantation, devices acquire alayer of host protein derived from plasma or interstitial

fluid.4 However, this protein layer may be responsible forphagocyte attraction and activation on the surface ofimplants.5 At the implant—tissue interface three host bloodproteins predominate: albumin, immunoglobulin G (IgG),and fibrinogen.6 The biomaterial surface and its physio-chemical properties are determinant for the adsorption anddenaturation of proteins such as fibrinogen.6,7 As a matterof fact, the fibrinogen has a key role in mediating the short-term accumulation of inflammatory cells on implanted bio-materials. The degree of inflammatory reaction is also influ-enced by the geometry of fibrinogen denaturation which isdeterminant to the exposure of pro-inflammatory epi-topes.4,5 Biomaterials are known agonists of the comple-ment system and leukocytes activation, both of them are theprelude of platelet activation and by consequence have arole in thrombosis. Besides inflammation, physiochemicalproperties of biomaterials also influence thrombosis that isa special kind of inflammation.8 Polyvinyl alcohol hydrogel(PVA) is a water-soluble synthetic polymer with an increas-ingly use in biomedical applications.9 It is produced by poly-merization of vinyl acetate to poly(vinyl) acetate followedby hydrolysis of polyvinyl acetate to polyvinyl alcohol. PVAmust be cross linked by physical or chemical methods inorder to be useful for a wide variety of biomedical applica-tions especially in the areas of medicine and pharmaceuti-cals sciences. A hydrogel can be defined as a hydrophiliccross-linked polymer (network) which swells when placedin water or biological fluids.9 However, it remains insolublein solution due to the presence of crosslinks. The PVA usehas been extended to biomedical applications including con-tact lenses, wound dressings, local drug delivery systems,and catheters.10–12 More recently, other biomedical applica-tions were tested under experimental conditions like vascu-lar grafts, artificial meniscus, intervertebral disk prosthesis,and also as a vitreous substitute.13–16 PVA was also beenused as a scaffold for biosynthetic cartilage.17,18 The basicconcept of tissue engineering consists in the seeding of cellsin a biodegradable scaffold impregnated or not with growthfactors and/or cytokines.19 The use of extra-embryonic tis-sues as stem cell reservoirs for tissue engineering offermany advantages over both embryonic and adult stem cellsources. Most significantly, the comparatively large volumeof extra-embryonic tissues and ease of physical manipula-tion theoretically increases the number of stem cells thatcan be isolated.20–23 These cells are capable of self-renewalwith sustained proliferation in vitro and can differentiateinto multiple mesodermal cells, including neuroglial-like

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4263

cells.20–23 The high plasticity and low immunogenicity ofthese cells turn them into a desirable form of cell therapyfor the injured tissues without requiring the use of immuno-suppressive drugs during the treatments.20–23 Throughthese immunosuppression properties of mesenchymal stemcells (MSCs), their association to vascular prosthetic surfacefabricated of different biomaterials can allow a faster bioin-tegration avoiding an exuberant local inflammatory reaction.For this reason, it was decided to evaluate the associationof MSCs isolated from the umbilical cord (UC) Wharton’sjelly with PVA in order to be used as a scaffold for vascularreconstruction in the sheep experimental model.

In this work, membranes of PVA were tested for biocom-patibility and hemocompatibility using sheep as an experi-mental model. This work aimed to test a biomaterial, thePVA with the future prospect of producing vascular grafts. Sofar, PVA has been tested for biocompatibility in rats and incrab-eating macaques and there is a lack of information ofthe behavior of this biomaterial in other species.14,24,25 Addi-tionally, MSCs from UC Wharton’s jelly were cultivated onPVA membranes in order to study the cytocompatibility andthe biointegration of the developed biomaterial associated tothis cellular system when used in vivo in pre-clinical trialswith sheep. To our knowledge, PVA gels prepared by freezingand thawing method were never been tested for biocompati-bility and hemocompatibility. With the aim of improving bio-compatibility and hemocompatibility, PVA was also tested incopolymerization with dextran (DX). DX is a polysaccharide,which is a biopolymer molecule with multiple effects inblood coagulation homeostasis including platelet activationinhibition,26 diminished fibrin polymerization,27 decreasedblood viscosity, and erythrocyte rouleaux formation,28 thatmight improve hemocompatibility if associated to other bio-materials like PVA.29

MATERIALS AND METHODS

Preparation of PVA membranesPVA cryogels were fabricated by a two-step gelation proce-dure involving a freeze/thaw cycle. A 20% PVA solutionwas prepared in distilled water at 80�C for several hoursto achieve homogeneity. In order to remove air bubbles,the former solution was immersed in an ultrasound bathfor 30 min. Solutions were poured into several polystyrenePetri dishes used as a mold of 500 mm thickness. Themembranes were subjected to three freeze/thaw cycleswhich leads to cryogel formation by physical reticulation.The freezing was done in a refrigerator at 228�C while thethawing process was done in an incubator at 25�C. Afterthe initial gels were obtained, 3 cm pieces of the gel werecut and further reinforced physically by immersing in asolution of 1M KOH and 1M Na2SO4 at room temperatureunder constant mixing for 1 h and afterward rinsing withdistilled water.

Preparation of PVA-dextran membranesFor the production of PVA membranes modified with DX(Sigma AldrichVR , molecular weight: 64,000–76,000 Da, St.Louis, USA) [Fig. 1(B)], two solutions were mixed, onesolution of PVA at 20% (w/v) with a solution of DX at 1%(w/v). The final mixture was made using the proportionDX to PVA at 10/90 (v/v). The DX solution was added tothe PVA solution when the PVA powder was completelydissolved. After the final solution was completed, it wasimmersed in an ultrasound bath for 30 min for removingair bubbles. Physical reticulation with three cycles wasused, followed by the annealing process for the productionof PVA/DX membranes. The PVA/DX membranes wereimmersed in a NaOH solution followed by the hydration indistilled water.

FIGURE 1. Monocultures of Human MSCs from Wharton’s jelly exhibiting a mesenchymal-like shape with a flat polygonal morphology (A)

(3100). SEM image of a MSC isolated from the Wharton’s jelly of the umbilical cord cultured over a PVA with dextran disk (B) (31500). MSCs

phenotype was confirmed by flow cytometry before and showed that over 95% of the cells were positive for the cell surface markers CD44,

CD73, CD90, and CD105 and less than 2% positive for CD14, CD19, CD31, CD34, CD45, and HLA-DR. [Color figure can be viewed in the online

issue, which is available at wileyonlinelibrary.com.]

4264 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

MSCs culture on PVA membranesMSCs from Wharton’s jelly UC matrix were purchased fromPromoCell GmbH (C-12971, lot-number: 8082606.7). TheMSCs were cultured and maintained in a humidified atmos-phere with 5% CO2 at 37�C. Mesenchymal Stem CellMedium, PromoCell (C-28010), was replaced every 48 h. At80% confluence, cells were harvested with 0.25% trypsinwith EDTA (GIBCO) and passed into a new flask for furtherexpansion. MSCs at a concentration of 104 cells/cm2 werecultured exhibiting an 80% confluence after 4 days onimmersed PVA membranes in culture medium [Fig. 1(A,B)].The phenotype of MSCs was assessed by PromoCell assay.Rigid quality control tests were performed for each lot ofPromoCell MSCs isolated from Wharton’s jelly of UC. MSCswere tested for cell morphology, adherence rate, and viabil-ity. Furthermore, each cell lot was characterized by flowcytometry analysis for a comprehensive panel of markers,such as platelet endothelial cell adhesion molecule-1(CD31), homing cell adhesion molecule (CD44), CD45, andEndoglin (CD105). The MSCs exhibited a mesenchymal-likeshape with a flat and polygonal morphology. During expan-sion, the cells became long spindle-shaped and colonizedthe whole culturing surface and PVA membranes [Fig.1(A,B)]. The MSCs phenotype was confirmed by flow cytom-etry before in vivo testing. Detection was performed withthe following antibodies and their respective isotypes (allfrom BioLegend unless stated otherwise): PE anti-humanCD105 (eBioScience); APC anti-human CD73; PE anti-humanCD90; PerCP/Cy5.5 anti-human CD45: FITC anti-humanCD34; PerCP/Cy5.5 anti-human CD14; Pacific Blue anti-human CD19; and pacific-blue anti-human HLA-DR. Chromo-some analysis on MSCs cell line from Wharton’s jelly beforein vivo application was carried out between passages 4 and5. When 80% confluence was reached, culture medium waschanged and supplemented with 4 lg/mL colcemid solution(stock solution, Cat. no. 15212-012, Gibco, NY). After 4 h,cells were collected and suspended in 8 mL of 0.075M KClsolution supplemented with bovine fetal serum. Then, thesuspension was incubated in 37�C for 35 min. After centrif-ugation (1500 rpm), 8 mL of the fixative methanol: glacialacetic acid at 6:1 was added and mixed together, and thecells were again centrifuged. After two rounds of fixation,two new rounds were performed with the fixative methanol:glacial acetic acid at 3:1. After the last centrifugation, thecell suspension was spread onto very well cleaned slides.Chromosome analysis was performed by one scorer on 20Giemsa-stained metaphases. Each cell was scored for chro-mosome number. Routine chromosome G-banding analysiswas also carried out for determination of the karyotype.The karyotype of the MSCs was determined and no struc-tural alterations were found demonstrating the absence ofneoplastic characteristics in these cells, as well as chromo-somal stability to the cell culture procedures.30 Intracellularfree Ca21 concentration ([Ca21]i) was measured in Fura-2-loaded MSCs cells by using dual wavelength spectrofluorom-etry as previously described.31 The measurements were per-formed on MSCs cultured on PVA-discs in order to correlatethe MSCs survival capacity in the presence of the PVA.

Results obtained from epifluorescence technique arereferred to measurements from MSCs which correspond to[Ca21]i from cells that did not begin the apoptosis process(data not shown). According to these results, it is reasona-ble to conclude that PVA is a viable substrate for undifferen-tiated MSCs vascular local delivery.

Scanning electron microscopyBefore scanning electron microscopy (SEM) analysis, theMSCs cultured on PVA disks [Fig. 1(B)] were first fixed with1.5% glutaraldehyde in 0.14M sodium cacodylate buffer (pH7.3) for 2 h at 4�C. Afterward, the PVA samples with andwithout the MSCs were dehydrated using graded ethanolsolutions from 60% to 100%, 5 min each, and subjected tocritical point drying. Finally, the samples were mounted onaluminum stubs using double-side adhesive tape and sput-ter coated with gold/palladium thin film, using the SPI Mod-ule Sputter Coater equipment for 100 s and with a 15 mAcurrent. The SEM/EDS exam was performed using a Highresolution (Schottky) Environmental Scanning ElectronMicroscope with X-Ray Microanalysis and Electron Backscat-tered Diffraction analysis: Quanta 400 FEG ESEM/EDAXGenesis X4M.

In vivo biocompatibility studiesExperimental groups. Thirty adult female white Merinosheep weighing �60 kg were randomly divided into fivegroups of six animals each. The animals were kept in strawyard (4 m 3 3 m) and were fed with standard chow andwater ad libitum. A diurnal 12 h light/12 h dark cycle wasused in this study. Adequate measures were taken to mini-mize pain and discomfort taking in account human end-points for animal suffering and distress during the in vivotesting. All procedures were performed with the approval ofthe Veterinary Authorities of Portugal in accordance withthe European Communities Council Directive of November1986 (86/609/EEC).

One group of animals (Group 1) was subcutaneouslyimplanted with PVA disks with 15.5 mm diameter; a secondgroup (Group 2) was subcutaneously implanted with PVAdisks covered with a confluent monoculture of human MSCsderived from Wharton’s jelly. The third group wasimplanted with PVA/DX membranes with the same diameter.As required by the ISO Standard 10993-6, a negative controlgroup (Group 4, sham) of six animals was included. Thepositive control group (Group 5) was subcutaneouslyimplanted with expanded polytetrafluoroethylene disk sam-ples (MAXIFLOTM, ePTFE Vascular Prosthesis, Vascutek,Scotland) with an equivalent area of biomaterial to the PVAmembranes implanted.

Surgical procedures. The animals were solids fastened for48 h and liquids were also removed 2 h before the surgicalprocedure. The anesthetic protocol consisted in xylazine(0.2 mg/kg, intravenous) as tranquilizer. For inducing anes-thesia, sodium thiopental was used at a dose of 15 mg/kgintravenous. The anesthetic maintenance was done with iso-flurane at 2% via endotracheal intubation carried by 100%

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4265

oxygen with a flow of 2 L/min. The wool was clipped inroughly square areas and the skin was aseptic prepared withiodopovidone solution. The incisions were made over thedorsal area, starting over the last ribs and ending just crani-ally to the sacrum. These 3 cm length incisions, three on theleft side and two on the right one, were made parallel and 4cm off the dorsal midline. A minimum distance of 3 cm waskept between incisions. The disks samples were insertedunder the skin, subcutaneously, distally to the midline. Subse-quently, the skin was closed with a SupramidVR USP1 sutureand the ewes were transferred to straw yards (4 m 3 3 m)after surgery. For the sham surgery, the surgical techniquewas performed exactly as previously described with theexception of the disk implantation. The skin sutures were notremoved in order to localize the correct implant site.

Retrieval of implants and histological evaluation. Theimplants and the surrounding tissue were collected after aperipheral infiltration with 2% lidocaine, at 1, 2, 4, 8, 16,and 32 weeks postimplantation. Five samples were collectedin each animal, randomly selected in each experimentalgroup at the previously reported temporal points. The sam-ples were fixed in 10% formalin, paraffin-embedded, cut in2 mm and stained with hematoxylin and eosin (HE) for his-tological evaluation.

The biological response parameters were assessed in theimplant/tissue interface with three high power fields (4003)by at least two experienced pathologists, for each sample andrecorded in an appropriated formulary. When there was adivergence of opinion, an agreed diagnosis was reached byusing a multihead microscope. Among the biological responseparameters, all were evaluated according to the ISO standard10993-6 (annex E) and included: the extent of fibrosis/fibrous capsule (layer in micrometers) and inflammation; thedegeneration as determined by changes in tissue morphol-ogy; the number and distribution from the material/tissueinterface of the inflammatory cell types, namely polymorpho-nuclear neutrophilic leucocytes (PMN), lymphocytes, plasmacells, eosinophils, macrophages, and multinucleated cells; thepresence, extent, and type of necrosis; other tissue altera-tions such as vascularization, fatty infiltration, and granulomaformation; the material parameters such as fragmentationand/or debris presence, form and location of remnants ofdegraded material. According to the scoring system based onthe mentioned parameters, a total value was obtained foreach animal and for each experimental group. The groupvalue was subtracted to the control group value and the testsample was considered as follows: Nonirritant (0.0 up to0.9), slight irritant (3.0 up to 8.9), moderate irritant (9.0 upto 15.0), and severe irritant (>15), to the tissue as comparedwith the control sample-irritant. Adverse histologicalresponses were documented by photomicrograph. The col-lected data was submitted to statistical analyses.

In vitro hemocompatibility studiesThe in vitro hemocompatibility studies followed the Stand-ard Practice for Assessment of Hemolytic Properties ofMaterials from the American Society for Testing and

Materials (ASTM F756-00, 2000), and comprehended theclassic hemolysis assay determination. According with thementioned standard, a material is considered nonhemolyticif the hemolytic index is inferior to 2%, slightly hemolytic ifbetween 2% and 5% and hemolytic if that value is over 5%.Three blood samples of healthy sheep and humans volun-teers were collected in sodium citrate tube. For each bloodsample, assays were made in duplicate. The PVA samplestested were: PVA, PVA plus 1% DX, PVA plus 10% DX, andas positive control, disks of expanded ePTFE (MAXIFLOTM

ePTFE Vascular Prosthesis, Vascutek, Scotland) were used.For the negative control, blood of sheep or human wasused. Before assay procedures, membranes were disinfectedby immersion in ethanol 70% (v/v) for 5 min followed bygenerous rinse with physiologic saline 0.9%. Both types ofblood (human and sheep) were diluted in phosphate buf-fered saline, pH 7.4, aiming to a concentration of 8 g/dL ofhemoglobin. Following blood dilution, the different biomate-rial samples and both controls were placed in contact withblood in six-well plates (in duplicate). The plates wereshacked every other hour and incubated at 37�C for 3 h. Atthe end of this period, the bloods were removed from theplates and centrifuged (Thermo Electron Corporation, JouanBr4i multifunction centrifuged) at 4000 rpm for 5 min at4�C. The supernatant of each centrifuged tube was removedwith the help of micropipette and transferred for a 96 wellsplate. Hemolysis was then spectrophotometrically (BioTekVR , Power wave XS) determined according to Ko et al.32

by means of a calibration curve (absorbance vs. hemoglo-bin) obtained as follows: a calibration curve plotting absorb-ance (at 540 nm) against hemoglobin was prepared bydiluting 1:100 the 8 g/dL hemoglobin initial suspension indistilled water to induce total hemolysis; with this dilution,a concentration of 80 mg/dL was obtained; by performingeight consecutive double dilutions with water, it was madea set of seven lysates standards with values between 800and 12.5 mg/L; the absorbance of each solution was read at540 nm, using water as blank.

Statistical analysesStatistical analysis was performed using the SPSS version19.0 (SPSS, Chicago, IL). Results are presented asmean6 SD. Multiple comparisons between groups were per-formed by one-way ANOVA supplemented with Tukey’s HSDpost hoc test. Differences were considered statistically signif-icant at p< 0.05.

RESULTS

The complications associated to medical devices are largelyassociated to material–tissue adverse reactions includingboth the effects of the implant on the host tissues, as wellas the effects of the host tissues on the implant. As a prom-ising material for small caliber vascular grafts, PVA wasevaluated in terms of its in vivo biocompatibility and hemo-compatibility. The effects of the MSCs isolated from the UCWharton’s jelly and DX in the improvement of PVA biocom-patibility and biointegration were also assessed. MSCs, as

4266 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

defined by the International Society for Cellular Therapy,are cells characterized by: (a) their capacity to adhere toplastic; (b) expression of specific surface markers, namely,CD73, CD90, and CD105, and no expression of CD14, CD19,CD34, CD45, and HLA-DR.33 The MSCs phenotype was con-firmed by flow cytometry before in vivo testing. As expectedfor MSCs, flow cytometry analysis showed that over 95% ofthe cells in the population were consistently positive for thecell surface markers CD44, CD73, CD90, and CD105 and lessthan 2% positive for CD14, CD19, CD31, CD34, CD45, andHLA-DR.

Considering the subcutaneous implants, the histologicalobservation showed that at week 1 postimplantation there

was a predominance of PMN and macrophage infiltrate atthe implant-tissue junction in both experimental group 1(PVA), group 2 (PVA plus MSCs), group 3 (PVA Dex 1%),and in the control group (ePTFE), suggesting an acute/sub-acute inflammatory reaction (Table I). During the first 2weeks, PMN and lymphocytes also prevailed over otherinflammatory cells, in the histological analysis performed tothe sham surgery group. In week 2 postimplantation, it wasobserved a significant increase in macrophages, plasmacells, and lymphocytes infiltration associated to a prolifera-tion of small blood vessels, compatible with a sub-acute/chronic inflammation. The persistence of the inflammatorystimuli in the group 1, group 2, group 3, and control group

TABLE I. Scoresa Attributed to the Level of PMN, Macrophage, Lymphocyte, Giant Cells, Fibrosis, and Neovascularization at 1,

2, 4, 8, 16, and 32 Weeks After Implantation

Weeksa,b

1 (n 5 5) 2 (n 5 5) 4 (n 5 5) 8 (n 5 5) 16 (n 5 5) 32 (n 5 5)

PMNGroup 1 (PVA) 2.8 (2,3) 2.4 (2,3) 1.4 (1,2) 1.4 (1,2) 1 (1,1) 0.2 (0,2)Group 2 (PVA 1 MSCs) 2.6 (1,4) 1.2 (0,2) 1 (0,2) 1 (0,2) 1 (0,1) 0 (0,0)Group 3 (PVA 1 dextran) 2.8 (2,3) 0.8 (0,1) 1 (1,1)c 1.4 (1,3) 1 (0,2) 0.2 (0,1)Group 4 (sham) 3.6 (3,4) 1.2 (0,2) 0.2 (0,1) 1.4 (1,3) 0 (0,0) 0.2 (0,1)Group 5 (ePTFE) 1.8 (1,2)e 1.6 (0,3) 0.2 (0,1)c 0 (0,0) 0.4 (0,1) 1 (0,2)

MacrophageGroup 1 (PVA) 1.2 (1,2) 2.6 (2,3) 3.6 (3,4) 3 (2,4) 3 (0,4) 2 (1,2)Group 2 (PVA 1 MSCs) 1.2 (1,2) 3.6 (3,4)c 3 (2,4) 2 (1,3)c 2 (1,3) 2 (1,3)Group 3 (PVA 1 dextran) 2.8 (2,3)c,d 2 (1,3)c,d 2 (1,2)c,d 2.6 (2,3)c 2.4 (1,3) 1.4 (1,2)c,d

Group 4 (sham) 0 (0,0)c,d,e 1 (1,1)d,e 0 (0,0)c,d,e 1 (0,3)d,e 0 (0,0) 2 (1,3)Group 5 (ePTFE) 1.2 (1,2)e,f 1.8 (1,3)d 1 (1,1)c,d,e,f 0 (0,0)c,f 0.6 (0,2) 0.8 (0,2)

LymphocyteGroup 1 (PVA) 2 (1,2) 1 (1,1) 2.2 (2,3) 2.4 (2,3) 3.2 (3,3) 2.2 (1,3)Group 2 (PVA 1 MSCs) 2.2 (1,4) 2 (1,4) 1.6 (1,2) 2 (1,2)c 1(1,2)c 2 (1,3)Group 3 (PVA 1 dextran) 2.6 (2,3) 2 (1,3)c 2 (1,3) 1.6 (1,2)d 2.4 (2,3)d 1.6 (1,2)c

Group 4 (sham) 1.6 (1,2) 2.2 (1,3) 1.4 (1,2) 2.6 (2,3)e 1.4 (1,2)d 2.2 (1,3)Group 5 (ePTFE) 2.2 (1,3) 1.4 (1,2) 1.2 (1,2) 1.2 (1,2)c,e 2 (1,3)c,e 2 (1,3)

Giant cellsGroup 1 (PVA) 0 (0,0) 1 (1,2) 1.4 (0,2) 1.4 (0,3) 1.4 (0,3) 0 (0,0)Group 2 (PVA 1 MSCs) 0.4 (0,1) 1.2 (0,2) 1.8 (1,3) 1 (0,2) 1 (0,3) 1 (0,3)Group 3 (PVA 1 dextran) 0.4 (0,2) 0.2 (0,1)c,d 0.4 (0,1)c,d 0.2 (0,1)c 0.4 (0,1) 0 (0,0)Group 4 (sham) 0 (0,0) 0 (0,0)c 0 (0,0)d 0 (0,0)c 0 (0,0) 0 (0,0)Group 5 (ePTFE) 0 (0,0) 0.8 (0,2) 0.4 (0,1)d 0 (0,0)c 0 (0,0) 1 (0,2)

FibrosisGroup 1 (PVA) 1.2 (1,2) 0.8 (0,1) 1.8 (1,2) 1 (1,1) 0 (0,0) 1 (1,1)Group 2 (PVA 1 MSCs) 1.8 (1,4) 0 (0,0) 2.4 (2,3) 2 (1,3)c 0 (0,1) 1 (0,1)Group 3 (PVA 1 dextran) 0.6 (0,1) 0.8 (0,1) 0.6 (0,1)c,d 0.8 (0,1)c,d 1.2 (1,2)c,d 1 (1,1)Group 4 (sham) 1 (1,1) 0.6 (0,1) 0 (0,0)c,d 0 (0,0)c,d,e 0 (0,0)e 1 (1,1)Group 5 (ePTFE) 2 (1,2) 3.2 (2,4)c,d,e,f 1.8 (1,2)e,f 1 (1,1)d,e 1.2 (1,2)c,d,e 2.4 (1,4)

NeovascularizationGroup 1 (PVA) 3 (3,3) 2.8 (2,4) 2.6 (2,3) 2.8 (2,3) 2.6 (2,3) 2.4 (2,3)Group 2 (PVA 1 MSCs) 3.2 (2,4) 3 (3,3) 2.2 (1,3) 3 (3,3) 2 (1,2) 2 (1,3)Group 3 (PVA 1 dextran) 2 (2,2) 1.8 (1,2) 1.8 (1,2) 1.6 (1,2)d 2.6 (2,3) 2 (1,3)Group 4 (sham) 2.4 (2,3) 2.8 (2,3)d 2 (1,3) 3.4 (3,4)c,e 2.4 (2,3) 2.4 (2,3)Group 5 (ePTFE) 2 (1,4) 3.2 (3,4)f 2.4 (2,3) 2.8 (2,3)f 3 (3,3) 3 (2,4)

Score: absent (0), rare (1), mild (2), moderate (3), and severe (4).

PVA, polyvinyl alcohol hydrogel; MSCs, mesenchymal stem cells; ePTFE, expanded polytetrafluorethylene.aThe number inside parenthesis represents the minimal and the maximal score observed in the event.bMean score of analysis.cp< 0.05 versus group 1.dp< 0.05 versus group 2.ep< 0.05 versus group 3.fp< 0.05 versus group 4.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4267

led to a change from acute to chronic inflammation (Fig. 2).However, in the sham surgery group, the macrophage num-ber showed a tendency to decrease (with the exception ofthe latest experimental point-32 weeks) probably due to thelack of antigenic stimuli. This chronic inflammationresponse was confined to the implantation site [Fig. 3(A)].

At week 4, the histological analysis revealed that the typeof cell infiltration consisted mainly in macrophage and lym-phocyte populations in all groups with the exception of thesham surgery group (Table I). From this time point forward,in groups 1 and 2, the multinucleated or giant cell numbersuffers an increase when compared with the previously timepoints. Concerning the other groups, giant cells were absentor rare during the 32 weeks. The type of cell infiltrate and theincreased neovascularization, always surrounding the implantarea were compatible with a chronic inflammation [Fig. 3(B)].

The foreign body reaction (FBR), consisting mainly ofmacrophages and/or foreign body giant cells infiltration atthe tissue–implant interface associated to the differentdegrees fibrosis (i.e., fibrous encapsulation), was observedat latter periods (16 and 32 weeks) of the experimentalprocess [Fig. 3(F,G)]. At the present work, the fibrosis grad-ing in the PVA containing membranes varied between theabsent (0) grade, to a mild (2) degree during the entireevaluated period of 32 weeks (Table I). Especially in the lat-est periods (16 and 32 weeks), the level of fibrosis wasreduced to absent (0) or rare (1). In general, the PVAgroups, particularly PVA-DX and PVA plus MSCs, presentedan inferior level of fibrosis (0–2) when compared with com-mercial available ePTFE (1–3).

By the histological analysis of the different experimentalgroups here reported, the number of macrophage remainedstable or showed a tendency to decrease, especially ingroup 2 where the PVA was associated to the cellular sys-tem (Table I). It was also observed a large number of neo-vasculatures, capillaries and venules, as well as fibrosisgranulation tissue at the material–tissue interface reinforc-ing the good biocompatibility of PVA and this in agreementwith our observations of new blood vessels around theimplants in the longest temporal points (16 and 32 weeks)for groups 1, 2, and 3, evidencing the good biocompatibilityof the biomaterial.

In this study, the neovascularization was similar to bothPVA groups (PVA and PVA plus MSCs) and the ePTFE group;especially in latter periods (16 and 32 weeks) (Table I). Inthe first weeks, the cells colonizing the implants weremostly neutrophils and macrophages. However, over time,macrophages became predominant over neutrophils, andfibroblasts were the main cell types within the implants,although blood vessels were restricted to the implant’speriphery. The FBR presented at tissue–PVA interface wasvery mild, particularly in group 3 (PVA-DX).

A more detailed analysis of the inflammation resultsover time showed us that PMN and lymphocyte prevailed inthe first weeks (1 and 2), indicative of an acute/sub-acuteinflammation. At the same time points, the observed necro-sis (data not showed) could also be attributed to the trau-matic injury of surgical procedure. Mild fibrosis and theabsence of giant cells during the several time points areindicative of the inexistence of FBR linked to the absence ofbiomaterial at surgical site. The observation of blood vesselsat the histologic samples evidenced moderate and stableamounts during the complete experimental period even atlate periods. For that reason, the presence of blood vesselsis probably connected to the intrinsic characteristics of sub-cutaneous connective tissue.

The results here presented showed that there is a tend-ency in the formation of a more intense fibrosis in late peri-ods as opposed to the first weeks (1 and 2 weeks). Therewere significant differences between PVA-DX, PVA, and PVAplus MSCs groups reinforcing the effect of DX in reducingFBR. However, for the sham surgery group, this value wasconsistently and significantly lower during the 32 weeksperiod which was obviously related with the absence of theimplant (Table I).

Analysis of the histological sections by light microscopyshowed the maturation of the granulation tissue, with aninflammatory infiltrate mainly composed by PMNs and lym-phocytes and no capsule at week 1 postimplantation, lead-ing the formation of a mature fibrous capsule surround theimplant and the presence of newly formed vessels at week32 postimplantation.

The scoring values for inflammation under the condi-tions for this study presented values as slightly irritant

FIGURE 2. Persistence of the inflammatory stimuli (PVA membrane) led to a change from acute to chronic inflammation. Predominance of PMN

(black arrow) at PVA–tissue interface, 1 week postimplantation (HE staining, 3200) (A), proliferation of giant cells (black arrow) and neovessels

at PVA–tissue interface (HE staining, 3200) (B), the presence of fibrotic capsule (black arrow) at PVA–tissue interface, 32 weeks postimplantation

(HE staining, 3200) (C). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

4268 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

(3.0–8.9) to nonirritant (0–2.9) (Table II). The biomaterials inthree groups were considered slight irritant to the tissue ascompared with the negative control sample (ePTFE). Even inthe sham surgery group (in spite of the absence of the bio-material), the achieved scoring value for inflammation was

1.3, despite being considered nonirritant confirms that thesurgical procedure also has a role in the inflammationprocess of biomaterial implantation. At week 4 and week 8postimplantation, a more intense inflammation in the experi-mental groups 1 and 2 was observed [Fig. 3(C,D)]. A marked

FIGURE 3. Chronic inflammation response (fibrotic capsule) locally confining the PVA membrane to the implantation site (HE staining, 340) (A).

Cell infiltrate and increased neovascularization, surrounding the implant area compatible with a chronic inflammation (HE staining, 3400) (B). At

4 (C) and 8 (D) weeks postimplantation, a more intense inflammation in the experimental groups 1 and 2 was observed (HE staining, (C) 3200,

(D) 3200). Histological image showing proliferation of fibroblasts, macrophages, and vascular endothelial cells (granulation tissue) and a

decrease in inflammatory cells such as PMN and lymphocytes at tissue–biomaterial interface, 4 weeks postimplantation (HE staining, 3100) (E).

Foreign body reaction consisting mainly of macrophages and/or foreign body giant cells infiltration at the tissue–implant interface associated to

the different degrees fibrosis at 16 (F) and 32 (G) weeks postimplantation (HE staining, (F) 3100 and (G) 3200). [Color figure can be viewed in

the online issue, which is available at wileyonlinelibrary.com.]

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4269

decrease in inflammation grading was detected in the severalexperimental and control groups at 8 weeks postimplantationonward (Table II). Although statistical significant differences(p<0.05) were verified between group 2 (PVA plus MSCs)and group 1 (PVA) at week 16 (Graphic 1). It was alsoobserved a statistically significant (p< 0.05) differencebetween group 3 and group 1/2 as concerns the inflamma-tory score at week 4 (Graphic 1). Also at week 16, the meaninflammatory score was statistically different (p< 0.05)between group 3 (PVA-DX) and group 1 (PVA) (Graphic 1).The overall mean inflammatory score was lower at group 2(PVA-MSCs), group 3 (PVA-DX), and group 4 (sham). How-ever, a statistically significant difference (p< 0.05) wasobserved between group 3 and group 1. This statistical evi-denced was also verified between group 4 and group 1. Dif-ferences were also verified among the sham surgery groupand group 3 (PVA-DX 1%) at latter experimental times. Thehistological results showed that none of the material testedinduced intense reaction with the host. The PVA implantscaused only a mild inflammatory reaction that was inferiorto commercially available products (ePTFE) and that inflam-matory reaction decreased along time (Table II).

A key component of biocompatibility of vascular graftsbiomaterials is the assessment of their interaction withblood or hemocompatibility. The in vitro hemocompatibilitystudies developed in this work were based in the in vitrohemolysis test. The interaction of PVA with blood was testedusing human and sheep blood. PVA was used isolated or inassociation with different concentrations (1% and 10%) oflow molecular weight DX. The hemolytic index values (%)were consistently higher for isolated PVA in either humans(0.028) or sheep (0.656) when compared with PVA-DX indifferent percentages of incorporation (1% and 10%), (TableIII). For positive control, it was used the ePTFE and theresults for the hemolytic index using human (0.016%) andsheep (0.014%) blood were lower than PVA and higherthan PVA-DX. PVA-DX presented the lowest value among thebiomaterials tested in the presented work; more specifically,PVA-DX 1% exhibited a value of 0.001% in sheep and0.002% in humans. In general, all the biomaterials tested inthe reported experimental work were considered nonhemo-lytic according with ASTM F756-00 standard.

DISCUSSION

As expected in the first 2 weeks, that PMN and lymphocyteprevailed over other inflammatory cells at tissue–biomateri-als interface indicative of an acute/sub-acute inflammation.From the first to the second week, the observed decrease inPMN values in all experimental groups can be explained bythe transition from the acute to sub-acute inflammatoryprocess (Table I). At the same time points, the observednecrosis (data not showed) could be attributed to the trau-matic injury of the surgical procedure and to the introduc-tion of biomaterial used for suturing. This event can alsocontribute to the inflammatory process apart from the bio-materials membranes. At later time points (from 4 weekonward), the type of cell infiltrate and the increasedT

AB

LE

II.

Mean

Infl

am

mato

ryS

co

resa

Over

Exp

eri

men

tal

Po

ints

for

the

Severa

lE

xp

eri

men

tal

Gro

up

s

Gro

up

s

Weeks

Mean

6S

Do

fA

llE

xp

eri

men

tal

Po

ints

1(n

55)

2(n

55)

4(n

55)

8(n

55)

16

(n5

5)

32

(n5

5)

Gro

up

1(P

VA

)(2

0.8

–20.4

)0.4

(21.8

–20)

1.8

(21.6

–11)

10.6

(20.2

–7.8

)12.4

(21.8

–12.4

)9.4

(13–1

7)

05.4

66

5.4

0G

rou

p2

(PV

A1

MS

Cs)

(18.6

–20.4

)0

(20.2

–20)

0.2

(20.6

–11)

9.6

(17.7

–0.8

)9.2

(10.6

–12.4

)0

b(1

4.2

–17)

03.1

66

4.4

0G

rou

p3

(PV

A1

dextr

an

)(2

1.4

–20.4

)1

(13.8

–20)

0(1

4.4

–11)

3.4

b,c

(15.2

–7.8

)7.4

(18.6

–12.4

)6.2

b(1

1–1

7)

03.0

06

8.8

1b

Gro

up

4(S

ham

surg

ery

)(1

5.4

–20.4

)0

d(1

1.2

–20)

0(1

2–1

1)

1d

(14.6

–7.8

)6.8

(6–1

2.4

)0

d(5

.8–1

7)

01.3

06

11.7

5b

Gro

up

5(e

PT

FE

)20.4

20

11

7.8

12.4

17

14.7

66

26.1

9

Th

en

um

ber

insi

de

the

pare

nth

esi

sre

pre

sen

tsth

ed

iffe

ren

ceb

etw

een

infl

am

mato

rysc

ore

of

con

tro

lg

rou

p(e

PT

FE

)an

dth

eo

ther

exp

eri

men

tal

gro

up

s.

PV

A,

po

lyvin

yl

alc

oh

ol

hyd

rog

el;

MS

Cs,

mese

nch

ym

al

stem

cells;

eP

TFE

,exp

an

ded

po

lyte

trafl

uo

reth

yle

ne.

n5

5.

aS

core

:N

on

irri

tan

t(0

.0u

pto

0.9

),sl

igh

tir

rita

nt

(3.0

up

to8.9

),m

od

era

teir

rita

nt

(9.0

up

to15.0

),an

dse

vere

irri

tan

t(>

15).

bp<

0.0

5vers

us

gro

up

1.

cp<

0.0

5vers

us

gro

up

2.

dp<

0.0

5vers

us

gro

up

3.

4270 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

neovascularization, always surrounding the implant area,were compatible with a chronic inflammation [Fig. 3(B)].Fibrosis is also a component of chronic inflammation due tothe growth factors secreted by macrophages acting in fibro-blast that leads to increase in fibrosis surrounding thebiomaterial.

Macrophages must be considered in the development ofimmune responses to synthetic biomaterials by presentingantigen to immune competent cells such as lymphocytesand plasma cells. The macrophage is probably the mostimportant cell in chronic inflammation because of the greatnumber of biologically active products expressed by thesecells. Among these, growth factors are responsible for thegrowth of fibroblasts (expressed histologically as fibrosis)and blood vessels observed in all groups at week 4 postim-plantation. Also these findings can be related to the forma-tion of granulation tissue, the hallmark of healinginflammation, consisting in the proliferation of fibroblasts,macrophages, and vascular endothelial cells and a decreasein inflammatory cells such as PMN and lymphocytes [Fig.3(E)]. The statistical significant differences (p< 0.05)observed between the three PVA groups from week 1 toweek 8 for the macrophage presence, being the group 3(PVA-DX) the experimental group that presented the lowestvalues on two of that time points (at week 2 and week 4;Table I). This finding probably is related to the cell limitingadhesion properties and consequently with limiting activa-tion of macrophage by DX.34 The FBR, consisting mainly ofmacrophages and/or foreign body giant cells infiltration atthe tissue–implant interface associated to the differentdegrees fibrosis (i.e., fibrous encapsulation), was observedat latter periods (16 and 32 weeks) of the experimentalprocess [Fig. 3(F,G)]. With biocompatible materials, the com-position of the FBR in the implantation site may be con-trolled by its surface properties, by the shape of the implant

and by the relationship between the surface area and thevolume of the implant.35,36 High surface-to-volume implantssuch as the PVA membranes would have higher counts ofmacrophages and foreign body giant cells in the tissue-implant surface. The great majority of biomaterials typicallyelicit a FBR, a special form of nonspecific chronic inflamma-tion. Fibrosis is considered the end-stage of FBR and con-sists in walling off the implant by a vascular andcollagenous fibrous capsule that is typically 50–200 lm inthickness.37,38 The low level of fibrosis observed by the useof PVA implants is indirectly related to a lesser degree ofmacrophage activation induced by PVA. Growth factors andcytokines (including platelet-derived growth factor (PDGF)and tumor growth factor (TGF-b)) known to be released byactivated macrophages at the onset of the foreign bodyresponse are potent mitogenics.39 In addition to their roleas potent mitogenics and chemotactics agents for myofibro-blast progenitors,40 PDGF-BB is associated with the earlystages of myofibroblast differentiation from progenitorcells.41 TGF-b is the principal mediator of myofibroblast dif-ferentiation in wound healing, inducing fibroblasts (and

TABLE III. Determined Values for Hemolysis Index Obtained

for PVA Isolated or Associated to Dextran

Biomaterial

SheepBlood—Hemolysis

Index (%)

HumanBlood—Hemolysis

Index (%)

PVA 0.656 0.028PVA 1 1% dextran 0.001 0.002PVA 1 10% dextran 0.005 0.017ePTFE 0.0014 0.016

Score: Nonhemolytic <2%, slightly hemolytic 2–5%, and

hemolytic >5%.

PVA, polyvinyl alcohol hydrogel; MSCs, mesenchymal stem cells;

ePTFE, expanded polytetrafluorethylene.

GRAPHIC 1. Inflammatory scores for the different experimental groups during the experimental period. Polyvinyl alcohol hydrogel (PVA), mesen-

chymal stem cells (MSCs), and expanded polytetrafluorethylene (ePTFE), statistical significant difference between experimental

groups. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4271

possibly other cell types) to differentiate into a-SM actin-expressing myofibroblasts with the capacity for contractionand extra cellular matrix synthesis.42 The fibrous wall con-fines the implant and consequently prevents it from inter-acting with the surroundings tissues which can bedeleterious to the biomedical devices if their function isbased in the permeability.

The form and topography of the surface of the biomate-rial determine the composition of the FBR. Relatively flatand smooth surfaces such as that found on breast prosthe-ses have a FBR that is composed of a layer of macrophagesone to two cells in thickness. Relatively rough surfaces suchas those found on the outer surfaces of expanded ePTFE orDacron vascular prostheses induce a FBR composed of mac-rophages and foreign body giant cells at the surface. Fabricmaterials generally have a surface response composed ofmacrophages and foreign body giant cells, with varyingdegrees of granulation tissue subjacent to the surfaceresponse. A material in a phagocytosable form (i.e., powderor particulate) may provoke a different degree of inflamma-tory response than the same material in a nonphagocytos-able form (i.e., film).43 Multinucleated giant cells (formed bythe fusion of monocytes and macrophages in an attempt tophagocytose the material with a size greater than the iso-lated cell) in the vicinity of a foreign body are generally con-sidered evidence of a more severe FBR. However, it is notuncommon to see very large foreign-body giant cells con-taining large numbers of nuclei on the surface of biomateri-als. In fact, the components of the FBR (giant cells andgranulation tissue) may persist at the tissue–implant inter-face for the lifetime of the implant.43,44 Only a few pub-lished studies address so far the in vivo biocompatibility ofPVA isolated or associated to other biomaterials. Mainly, invitro studies have been performed for assessing the biocom-patibility which has obvious limitations in extrapolating theresults to more complex in vivo systems. Noguchi et al.tested the biocompatibility of PVA as an artificial articularcartilage in the intra-articular and intramuscular environ-ment.45 Tissue reactions of cartilage, bone, synovium, andmuscle to PVA were histologically analyzed. In the tissues inwhich PVA was implanted, local inflammatory reaction wasvery slight, confirming the results here described of slightirritation to the surrounding tissues in the experimentalgroups 1, 2, 3, and sham. Seo et al. also studied PVA in com-bination with gelatin for biocompatibility by implantation inmuscular tissue of rabbits, confirming that PVA induces lessaccumulation of inflammatory cells when compared withthe control group consisting in polyurethanes.46 Burczaket al. used PVA to encapsulate Langerhans islets in macro-bags to make a hybrid-type artificial pancreas.47 The cellularenzyme activity in the implant-encapsulating tissue wasmeasured, regarding the acid and alkaline phosphatases.This studied intended to evaluate the activity of the cellsinvolved in the inflammatory response to the long-termmacrocapsule implantation. In the long-term experimentaltime points (133 days) of the work of Burczak et al.,25 itwas evidenced the maximum activity of acidic phosphatasethat is normally linked to a greater activity of macrophage.

By the histological analysis of the different experimentalgroups here reported, the number of macrophage remainedstable or showed a tendency to decrease over time, espe-cially in group 2 (PVA plus MSCs) where the PVA was asso-ciated to the cellular system (Table I). MSCs from theWharton’s jelly of the UC grown in aggregates that bettermimic tissue environment have produced a secretome richin trophic factors, such as HGF, TGF-b, G-CSF, VEGF-A, FGF-2, KGF, and IL-6 that promote wound healing reactions, asdemonstrated both in vitro by vasculogenesis, mitogenic,and chemotactic assays and in vivo, using a chemotaxisassay where MSCs were shown to recruit surrounding bonemarrow MSCs known to be directly involved in tissue regen-eration. The difference between our results and the resultspublished by Burczak et al.25 can be explained by the use ofdistinct cross-linking processes that can influence the num-ber of nucleophilic hydroxyl groups in the PVA macromole-cule which reacts with the C3 component, which activatesthe alternative pathway of the complement system and isalso responsible for the adhesion and activation of macro-phage among other inflammatory cells.48 It was alsoobserved a large number of neo-vasculatures, capillariesand venules, as well as fibrosis granulation tissue at thematerial–tissue interface reinforcing the good biocompatibil-ity of PVA and this in agreement with our observations ofnew blood vessels around the implants in the longest tem-poral points (16 and 32 weeks) for groups 1, 2, and 3, evi-dencing the good biocompatibility of the biomaterial.Cytotoxicity or cytocompatibility of PVA also was evaluatedby this study using L-929 fibroblasts cell culture, the per-centages of cell growth inhibition of samples were actuallyhigher in the control group suggesting that the PVA has asuperior cytocompatibility to polyurethane and low densitypolyethylene.46 Even associated to calcium biphosphate(BCP) and used as scaffold for bone regeneration, PVA hasproven to have no negative effects on cells growth and pro-liferation, and bone marrow derived stem cells possessed afavorable spreading morphology on the BCP/PVA scaffoldsurface.49

The FBR presented at tissue–PVA interface was verymild, particularly in group 3 (PVA with DX). The copolymer-ization of PVA with biological polymers such as polysaccha-rides has been reported in the literature due to theincreasing effect on biocompatibility of this association.50

DX has been tested for biocompatibility (by subcutaneousimplantation) in different percentage of incorporations. Inhigher values (75 and 100%) of added DX, the number ofmacrophage and lymphocyte where absent or reduced withincreased when compared with the control group. Suggest-ing that DX reduces FBR by limiting cell adhesion, spread-ing, and consequently activation.51,52 This observationssupport our results (Table I), that showed a statistical sig-nificant difference (p<0.05) for macrophage countingbetween PVA and PVA-DX groups during the entire experi-mental period with exception of the week 16. Mild fibrosisand the absence of giant cells during the several time pointsare indicative of the inexistence of FBR linked to theabsence of biomaterial at surgical site. The observation of

4272 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

blood vessels at the histologic samples evidenced moderateand stable amounts during the complete experimentalperiod even at late periods. For that reason, the presence ofblood vessels is probably connected to the intrinsic charac-teristics of subcutaneous connective tissue.

Generally, fibrosis (i.e., fibrous encapsulation) surroundsthe biomaterial, isolating the implant and foreign-body reac-tion from the local tissue environment.43 Our resultsshowed that there is a tendency in the formation of a moreintense fibrosis in late periods as opposed to the first weeks(1 and 2 weeks). There were significant differences betweenseveral groups more specifically between PVA-DX, PVA, andPVA plus MSCs groups reinforcing the effect of DX in reduc-ing FBR. However, for the sham surgery group, this valuewas consistently and significantly lower during the 32weeks period which was obviously related with the absenceof the implant (Table I). The thickness of the fibrous capsulearound implants placed subcutaneously has been used as ameasure of the biocompatibility of materials because fibro-sis and multinucleated giant cells are landmarks of FBR.Hence, overall, from the present work, a classification ofPVA as a biocompatible material may be drawn. However, itis important to note that materials yielding acceptable tis-sue compatibility in one implantation site might yield unfav-orable results in another site.43 Analysis of the histologicalsections by light microscopy showed the maturation of thegranulation tissue, with an inflammatory infiltrate mainlycomposed by PMNs and lymphocytes and no capsule atweek 1 postimplantation, leading the formation of a maturefibrous capsule surround the implant and the presence ofnewly formed vessels at week 32 postimplantation.

Histological results showed that the inflammatoryresponse to PVA is inferior or similar to that of commer-cially available products like ePTFE (Table II).The statisticalsignificant inferior difference observed for PVA-DX can beexplained by the positive influence of polysaccharides onbiocompatibility due to their anti-inflammatory effects.52–54

The results obtained until now with PVA (isolated or incombination with other polymers) implants produced onlymild inflammation which is in according with our observa-tions.55 The surgical procedure and the associated traumacan also be responsible for some part of the magnitude ofthat inflammation process rather than PVA itself as demon-strated by sham surgery group inflammatory scoring value(1.3). A marked decrease in inflammation grading wasdetected in the several experimental and control groups atweek 8 postimplantation onward (Table II). This tendencywas most pronounced at group 2 where the biomaterialwas associated to the MSCs. This result was probably due tothe immunomodulatory properties of MSCs, allowing afaster biointegration of the biomaterial and avoiding anexuberant local inflammatory reaction.56,57 These immuno-modulatory properties include the suppression of inflamma-tory cytokines and the induction of T cells with regulatoryor suppressive phenotypes.58 These actions were moreeffective at latter stages of inflammation were this cells pre-dominate. The histological results showed that none of thematerial tested induced intense reaction with the host.

Thus, to use PVA as a biomedical scaffold for artificial bloodvessels additional compatibility studies are required, inwhich the biomaterial is subjected to the existing environ-ment in the grafting site. The use of large animal models forbiocompatibility assessment of biomaterials used in vascularsurgery research has several advantages over laboratoryanimals: sheep and pigs presented a coagulation system tohumans and there is availability of large vessels that can beused in the implantation of artificial graft.59,60

The PVA membranes with DX had a marked decrease inthe hemolytic index which could be attributed to the cyto-protective effects of DX in erythrocytes. DX is a glycosamino-glycan that has been used by several authors has apreventive substance of hemolysis; however the mechanismthat produces that effect is poorly understood.61 The effectsof low molecular weight DX in erythrocytes are related tothe decrease of erythrocyte aggregation and blood viscositywhich can affect the interaction of those cells with biomateri-als.61 This polysaccharide also affects the membrane stabilityand creates a negative electrical charge on the surface, creat-ing repulsive forces between erythrocytes, decreasing aggre-gation, and hemolysis.62,63 During the in vivo trials periodwith PVA or ePTFE membranes, the collected samples didnot presented biodegradation of the vascular drafts. The ani-mals implanted with PVA disks showed an absence of localor metastatic tumorigenesis and local or system signs ofinfection as well. PVA is described in the literature as a bio-degradable polymer. Although, the biodegradation process, sofar described, is a microbial mediated process which pre-cludes their degradation by animal cellular enzymatic appara-tus.64 Neoplasms associated to therapeutic clinical implantsin humans and animals are rare and the large majority isdiagnosed as sarcomas associated to metal implants.65 TheFBR induced by the PVA might stimulate the formation of afibrous capsule which is conductive to a neoplastic processbut the mechanism is still poorly understood.65

However, until now for PVA isolated or with copolymer-ization, no episode of neoplasm has been reported in theliterature.

The association of different polymers with natural orartificial polysaccharides has been a current trend of investi-gation in biomaterials science. Their increasing use isrelated with the immunomodulatory, anticoagulant, andanti-inflammatories properties that lead to a better biocom-patibility of the resulting association. To our knowledge,PVA has never been tested for biocompatibility and hemo-compatibility in association with DX. The presented resultssupport the use of this copolymerization as a vascular graftin a pre-clinical animal model and also the association withMSCs from the Wharton’s jelly of the UC improves the bioin-tegration of the biomaterial modulating the local inflamma-tory reaction.

CONCLUSIONS

PVA vascular graft was demonstrated to be a biomaterialwith potential to be used in vascular reconstruction. Partic-ularly, the in vivo testing showed that PVA, PVA with DX,

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4273

and PVA covered by human MSCs isolated from the Whar-ton’s jelly of the UC are just slight irritant material to thesurrounding tissues according to the ISO standard 10993-6(annex E). Consequently, a biocompatible material that canbe used in the near future is considered as a vascular graftalso confirmed by an excellent blood–material interactionevaluated by hemocompatibility studies followed the Stand-ard Practice for Assessment of Hemolytic Properties ofMaterials from the American Society for Testing and Materi-als (ASTM F756-00, 2000), including the hemolysis assayclassic hemolysis determination. The results obtained by thein vitro blood–material interactions envisaged a nonhemo-lytic index biomaterial that needs further in vivo testing toconfirm that promising results.

ACKNOWLEDGMENTS

The authors thank ZEA—Sociedade Unipessoal, Lda, for thehelp in handling the animals during the experimental periods.

REFERENCES

1. World Health Organization. Cardiovascular Disease: Global Atlas

on Cardiovascular Disease Prevention and Control. Geneva, Swit-

zerland: Global Atlas on Cardiovascular Disease Prevention and

Control; 2012.

2. Cleary MA, Geiger E, Grady C, Best C, Naito Y, Breuer C. Vascular

tissue engineering: The next generation. Trends Mol Med 2012;

18:394–404.

3. Lyman DJ, Murray-Wijelath J, Ambrad-Chalela E, Wijelath ES.

Vascular graft healing. II. FTIR analysis of polyester graft samples

from implanted bi-grafts. J Biomed Mater Res 2001;58:221–237.

4. Tang L, Hu W. Molecular determinants of biocompatibility. Expert

Rev Med Dev 2005;2:493–500.

5. Hu WJ, Eaton JW, Ugarova TP, Tang L. Molecular basis of

biomaterial-mediated foreign body reactions. Blood 2001;98:1231–

1238.

6. Tang L, Eaton JW. Fibrin(ogen) mediates acute inflammatory

responses to biomaterials. J Exp Med 1993;178:2147–2156.

7. Tang L, Ugarova TP, Plow EF, Eaton JW. Molecular determinants

of acute inflammatory responses to biomaterials. J Clin Invest

1996;97:1329–1334.

8. Gorbet MB, Sefton MV. Biomaterial-associated thrombosis: Roles

of coagulation factors, complement, platelets and leukocytes. Bio-

materials 2004;25:5681–5703.

9. Hassan C, Peppas N. Structure and applications of poly(vinyl alco-

hol) hydrogels produced by conventional crosslinking or by freez-

ing/thawing methods biopolymers. In: Abe A, Albertsson AC,

editors. PVA Hydrogels, Anionic Polymerisation Nanocomposites.

Berlin: Springer; 2000. p 37–65.

10. Kokabi M, Sirousazar M, Hassan ZM. PVA-clay nanocomposite

hydrogels for wound dressing. Eur Polym J 2007;43:773–781.

11. Hyon SH, Cha WI, Ikada Y, Kita M, Ogura Y, Honda Y. Poly(vinyl

alcohol) hydrogels as soft contact lens material. J Biomater Sci

Polym Ed 1994;5:397–406.

12. Wang M, Li Y, Wu J, Xu F, Zuo Y, Jansen JA. In vitro and in vivo

study to the biocompatibility and biodegradation of hydroxyapa-

tite/poly(vinyl alcohol)/gelatin composite. J Biomed Mater Res

Part A 2008;85A:418–426.

13. Kobayashi M, Chang YS, Oka M. A two year in vivo study of poly-

vinyl alcohol-hydrogel (PVA-H) artificial meniscus. Biomaterials

2005;26:3243–3248.

14. Maruoka S, Matsuura T, Kawasaki K, Okamoto M, Yoshiaki H,

Kodama M, Sugiyama M, Annaka M. Biocompatibility of polyviny-

lalcohol gel as a vitreous substitute. Curr Eye Res 2006;31:599–

606.

15. Chaouat M, Le Visage C, Baille WE, Escoubet B, Chaubet F,

Mateescu MA, Letourneur D. A novel cross-linked poly(vinyl alco-

hol) (PVA) for vascular grafts. Adv Funct Mater 2008;18:2855–2861.

16. Joshi A, Fussell G, Thomas J, Hsuan A, Lowman A, Karduna A,

Vresilovic E, Marcolongo M. Functional compressive mechanics

of a PVA/PVP nucleus pulposus replacement. Biomaterials 2006;

27:176–184.

17. Bichara DA, Zhao X, Hwang NS, Bodugoz-Senturk H, Yaremchuk

MJ, Randolph MA, Muratoglu OK. Porous poly(vinyl alcohol)-algi-

nate gel hybrid construct for neocartilage formation using human

nasoseptal cells. J Surg Res 2010;163:331–336.

18. Bichara DA, Zao X, Bodugoz-Senturk H, Ballyns FP, Oral E,

Randolph MA, Bonassar LJ, Gill TJ, Muratoglu OK. Porous poly(-

vinyl alcohol)-hydrogel matrix-engineered biosynthetic cartilage.

Tissue Eng Part A 2011;17:301–309.

19. van Blitterswijk CA, Moroni L, Rouwkema J, Siddappa R, Sohier

J. Tissue engineering—An introduction. In: van Clemens B, Peter

T, Anders L, Jeffrey H, David FW, Ranieri C, Joost DdB, J�erome S,

editors. Tissue Engineering. Burlington: Academic Press; 2008. p

xii–xxxvi.

20. Marcus AJ, Woodbury D. Fetal stem cells from extra-embryonic

tissues: Do not discard. J Cell Mol Med 2008;12:730–742.

21. Fu YS, Cheng YC, Lin MYA, Cheng H, Chu PM, Chou SC, Shih YH,

Ko MH, Sung MS. Conversion of human umbilical cord mesen-

chymal stem cells in Wharton’s jelly to dopaminergic neurons in

vitro: Potential therapeutic application for parkinsonism. Stem

Cells 2006;24:115–124.

22. Wang HS, Hung SC, Peng ST, Huang CC, Wei HM, Guo YJ, Fu

YS, Lai MC, Chen CC. Mesenchymal stem cells in the Wharton’s

jelly of the human umbilical cord. Stem Cells 2004;22:1330–1337.

23. Maur�ıcio AC, G€artner A, Armada-da-Silva P, Amado S, Pereira T,

Veloso AP, Varej~ao A, Lu�ıs AL, Geuna S. Cellular systems and

biomaterials for nerve regeneration in neurotmesis injuries. In:

Pignatello R, editor. Biomaterials Applications for Nanomedicine.

Rijeka, Croatia, 2011. p 415–440.

24. Wang LC, Chen XG, Zhong DY, Xu QC. Study on poly(vinyl alco-

hol)/carboxymethyl-chitosan blend film as local drug delivery sys-

tem. J Mater Sci: Mater Med 2007;18:1125–1133.

25. Burczak K, Gamian E, Kochman A. Long-term in vivo performance

and biocompatibility of poly(vinyl alcohol) hydrogel macrocap-

sules for hybrid-type artificial pancreas. Biomaterials 1996;17:

2351–2356.

26. Zeerleder S, Mauron T, L€ammle B, Wuillemin WA. Effect of low-

molecular weight dextran sulfate on coagulation and platelet

function tests. Thromb Res 2002;105:441–446.

27. Abir F, Barkhordarian S, Sumpio BE. Efficacy of dextran solutions

in vascular surgery. Vasc Endovascular Surg 2004;38:483–491.

28. Neu B, Wenby R, Meiselman HJ. Effects of dextran molecular

weight on red blood cell aggregation. Biophys J 2008;95:3059–

3065.

29. Fasl H, Stana J, Stropnik D, Strnad S, Stana-Kleinschek K,

Ribitsch V. Improvement of the hemocompatibility of PET surfa-

ces using different sulphated polysaccharides as coating materi-

als. Biomacromolecules 2010;11:377–381.

30. G€artner A, Pereira T, Alves MG, Armada-da-Silva PAS, Amorim I,

Gomes R, Ribeiro J, Franca ML, Lopes C, Carvalho RA, Socorro S,

Oliveira PF, Porto B, Sousa R, Bombaci A, Ronchi G, Fregnan F,

Varej~ao ASP, Lu�ıs AL, Geuna S, Maur�ıcio AC. Use of poly(D,L-lac-

tide—caprolactone) membranes and mesenchymal stem cells

from the Wharton’s jelly of the umbilical cord for promoting

nerve regeneration in axonotmesis: In vitro and in vivo analysis.

Differentiation 2012;84:355–365.

31. Amado S, Sim~oes MJ, Armada da SP, Lu�ıs AL, Shirosaki Y, Lopes

MA, Santos JD, Fregnan F, Gambarotta G, Raimondo S, Fornaro

M, Veloso AP, Varej~ao ASP, Maur�ıcio AC, Geuna S. Use of hybrid

chitosan membranes and N1E-115 cells for promoting nerve

regeneration in an axonotmesis rat model. Biomaterials 2008;29:

4409–4419.

32. Ko FN, Hsiao G, Kuo YH. Protection of oxidative hemolysis by

demethyldiisoeugenol in normal and thalassemic red blood cells.

Free Radical Biol Med 1997;22:215–222.

33. Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini FC,

Krause DS, Deans RJ, Keating A, Prockop DJ, Horwitz EM. Mini-

mal criteria for defining multipotent mesenchymal stromal cells.

The International Society for Cellular Therapy position statement.

Cytotherapy 2006;8:315–317.

4274 ALEXANDRE ET AL. BIOCOMPATIBILITY AND HEMOCOMPATIBILITY OF PVA HYDROGEL

34. Massia SP, Stark J, Letbetter DS. Surface-immobilized dextran

limits cell adhesion and spreading. Biomaterials 2000;21:2253–

2261.

35. Rae T. The macrophage response to implant materials with spe-

cial reference to those used in orthopedics. Crit Rev Biocompat

1986;2:97–126.

36. Greisler H. Macrophage-biomaterial interactions with bioresorb-

able vascular prostheses. ASAIO Trans 1988;34:1051–1059.

37. Williams GT, Williams WJ. Granulomatous inflammation—A

review. J Clin Pathol 1983;36:723–733.

38. Kovacs EJ. Fibrogenic cytokines: The role of immune mediators

in the development of scar tissue. Immunol Today 1991;12:17–23.

39. Luttikhuizen DT, Harmsen MC, Van Luyn MAJ. Cellular and

molecular dynamics in the foreign body reaction. Tissue Eng

2006;12:1955–1970.

40. Lindahl P, Christer. Not all myofibroblasts are alike: Revisiting the

role of PDGF-A and PDGF-B using PDGF-targeted mice. Curr Opin

Nephrol Hypertens 1998;7:21–26.

41. Oh SJ, Kurz H, Christ B, Wilting J. Platelet-derived growth factor-

B induces transformation of fibrocytes into spindle-shaped myofi-

broblasts in vivo. Histochemistry 1998;109:349–357.

42. Leask A, Abraham DJ. TGF-B signaling and the fibrotic response.

FASEB J 2004;18:816–827.

43. Ratner BD, Schoen FJ. Chapter II. 3.2—The concept and assess-

ment of biocompatibility. In: Buddy DR, Hoffman AS, Frederick

JSaJ, Lemons JE, editors. Biomaterials Science, 3rd ed. Oxford,

UK: Academic Press; 2013. p 588–592.

44. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to

biomaterials. Semin Immunol 2008;20:86–100.

45. Noguchi T, Yamamuro T, Oka M, Kumar P, Kotoura Y, Hyonyt SH,

Ikadat Y. Poly(vinyl alcohol) hydrogel as an artificial articular carti-

lage: Evaluation of biocompatibility. J App Biomater 1991;2:101–107.

46. Seo KH, You SJ, Chun HJ, Who CHK, Lim Y. In vitro and in vivo

biocompatibility of x-ray crosslinked gelatin-poly(vinyl alcohol)

hydrogels. Tissue Eng Regen Med 2009;6:414–418.

47. Burczak K, Gamian E, Kochman A. Long-term in vivo performance

and biocompatibility of poly(vinyl alcohol) hydrogel macrocap-

sules for hybrid-type artificial pancreas. Biomaterials 1996;17:

2351–2356.

48. Sawada K, Shimoyama T, Malchesky PS, Goldcamp JB,

Omokawa S. Evaluation of a relationship between polymer bulk

hydroxyl and surface oxygen content and in vitro serum-material

interaction. J Biomed Mater Res 1993;27:547–555.

49. Nie L, Chen D, Suo J, Zou P, Feng S, Yang Q, Yang S, Ye S. Phys-

icochemical characterization and biocompatibility in vitro of

biphasic calcium phosphate/polyvinyl alcohol scaffolds prepared

by freeze-drying method for bone tissue engineering applications.

Colloids Surf B 2012;100:169–176.

50. Alhosseini SN, Moztarzadeh F, Mozafari M, Asgari S, Dodel M,

Samadikuchaksaraei A, Kargozar S, Jalali N. Synthesis and char-

acterization of electrospun polyvinyl alcohol nanofibrous scaf-

folds modified by blending with chitosan for neural tissue

engineering. Int J Nanomed 2012;7:25–34.

51. Cad�ee JA, van Luyn MJA, Brouwer LA, Plantinga JA, van

Wachem PB, de Groot CJ, den Otter W, Hennink WE. In vivo

biocompatibility of dextran-based hydrogels. J Biomed Mater Res

2000;50:397–404.

52. Abed A, Assoul N, Ba M, Derkaoui SM, Portes P, Louedec L,

Flaud P, Bataille I, Letourneur D, Meddahi-Pell�e A. Influence of

polysaccharide composition on the biocompatibility of pullulan/

dextran-based hydrogels. J Biomed Mater Res Part A 2011;96A:

535–542.

53. Li H, Chen C, Zhang S, Jiang J, Tao H, Xu J, Sun J, Zhong W,

Chen S. The use of layer by layer self-assembled coatings of hyal-

uronic acid and cationized gelatin to improve the biocompatibility

of poly(ethylene terephthalate) artificial ligaments for reconstruc-

tion of the anterior cruciate ligament. Acta Biomater 2012;8:4007–

4019.

54. Alauzun JG, Young S, D’Souza R, Liu L, Brook MA, Sheardown

HD. Biocompatible, hyaluronic acid modified silicone elastomers.

Biomaterials 2010;31:3471–3478.

55. Elakkiya T, Sheeja R, Ramadhar K, Natarajan TS. Biocompatibility

studies of electrospun nanofibrous membrane of PLLA-PVA

blend. J Appl Polym Sci 2013;128:2840–2846.

56. Koh CJ, Atala A. Tissue engineering, stem cells, and cloning: Cur-

rent concepts and future trends. In: Keating A, Dicke K, Gorin N,

Weber R, Graf H, editors. Regenerative and Cell Therapy. Berlin:

Springer; 2005. p 35–67.

57. Martino S, D’Angelo F, Armentano I, Kenny JM, Orlacchio A.

Stem cell-biomaterial interactions for regenerative medicine. Bio-

technol Adv 2012;30:338–351.

58. Yagi H, Soto-Gutierrez A, Parekkada B, Kitagawa Y, Tompkins RG,

Kobayashi N, Yarmush ML. Mesenchymal stem cells: Mecha-

nisms of immunomodulation and homing. Cell Transplant 2010;

19:667–679.

59. Cascone MG, Maltinti S, Barbani N, Laus M. Effect of chitosan

and dextran on the properties of poly(vinyl alcohol) hydrogels. J

Mater Sci: Mater Med 1999;10:431–435.

60. Ueberrueck T, Tautenhahn J, Meyer L, Kaufmann O, Lippert H,

Gastinger I, Wahlers T. Comparison of the ovine and porcine ani-

mal models for biocompatibility testing of vascular prostheses. J

Surg Res 2005;124:305–311.

61. Eisenberg S. The effect of low molecular weight dextran on the

viscosity and suspension characteristics of blood. Am J Med Sci

1969;257:336–343.

62. Fernandes JC, Eaton P, Nascimento H, L€u B, Rocha S, Vitorino R,

Amado F, Gomes J, Santos-Silva A, Pintado ME, Malcata FX.

Effects of chitooligosaccharides on human red blood cell mor-

phology and membrane protein structure. Biomacromolecules

2008;9:3346–3352.

63. Fernandes HP, Cesar CL, Barjas-Castro MdL. Electrical properties

of the red blood cell membrane and immunohematological inves-

tigation. Rev Bras Hematol Hemoter 2011;33:297–301.

64. Chiellini E, Corti A, D’Antone S, Solaro R. Biodegradation of

poly(vinyl alcohol) based materials. Prog Polym Sci 2003;28:963–

1014.

65. Schoen FJ. Chapter II. 2.7—Tumors associated with biomaterials

and implants. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE,

editors. Biomaterials Science, 3rd ed. Oxford, UK: Academic

Press; 2013. p 558–565.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | DEC 2014 VOL 102A, ISSUE 12 4275