Journal of Surgical Research 163, 331–336 (2010)doi:10.1016/j.jss.2010.03.070

ASSOCIATION FOR ACADEMIC SURGERY

Porous Poly(vinyl alcohol)-Alginate Gel Hybrid Construct for Neocartilage

Formation Using Human Nasoseptal Cells

David A. Bichara, M.D.,* Xing Zhao, M.D.,* Nathaniel S. Hwang, Ph.D.,† Hatice Bodugoz-Senturk, Ph.D.,‡Michael J. Yaremchuk, M.D.,* Mark A. Randolph, M.A.S.,*,1 and Orhun K. Muratoglu, Ph.D.‡

*Plastic Surgery Research Laboratory, Division of Plastic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston,Massachusetts; †Koch Institute for Innovative Cancer Research, Massachusetts Institute of Technology, Cambridge, Massachusetts;and ‡Harris Orthopaedic Biomechanics and Biomaterials Laboratory, Massachusetts General Hospital, Harvard Medical School,

Boston, Massachusetts

Submitted for publication January 8, 2010

Background. Limited options exist for the restora-tion of craniofacial cartilage. Autologous tissue or po-rous polyethylene is currently used for nasal andauricular reconstruction. Both options are associatedwith drawbacks, including donor site morbidity andimplant extrusion. Poly(vinyl alcohol) (PVA) is a non-degradable flexible biocompatible polymer than canbe engineered to mimic the properties of cartilage.The goal of this study was to engineer a biosynthetichybrid construct using a combination of PVA-alginatehydrogels and human nasal septum chondrocytes.

Materials and Methods. Chondrocytes isolated fromhuman nasal septum cartilage were expanded andmixed with 2% sodium alginate hydrogel. Thechondrocyte-alginate mix was injected into a non-degradable porous PVA hydrogel, creating biosyn-thetic constructs. A group of these constructs wereimplanted into the subcutaneous environment ofnude mice, while the other group was cultured ina spinner flask bioreactor system for 10 d and then im-planted. After 6 wk in vivo, the histologic, biochemical,and biomechanical properties were examined.

Results. Histological analysis demonstrated sul-fated glycosaminoglycans and deposition of collagentype II in constructs from both groups. Constructs cul-tured in the bioreactor system prior in vivo implanta-tion demonstrated higher levels of DNA,glycosaminoglycans, and hydroxyproline. An increaseof 22% in the compressive strength of the engineeredconstructs exposed to the bioreactor was also ob-served.

1 To whom correspondence and reprint requests should be ad-dressed at Massachusetts General Hospital, Room WAC 435, 15 Park-man Street, Boston, MA 02114. E-mail: marandolph@partners.org.

331

Conclusion. A novel porous PVA-alginate gel hybridwas used to successfully engineer human cartilagein vivo. A 10-d period of bioreactor culturing increasedlevels of DNA, glycosaminoglycans, hydroxyproline,and the compressive modulus of the constructs. � 2010

Elsevier Inc. All rights reserved.

Key Words: poly(vinyl alcohol); hydrogel; cartilage;nasoseptal chondrocytes; tissue engineering.

INTRODUCTION

Reconstruction and restoration of the cartilaginousstructures of the craniofacial region continues to bea challenge for surgeons. The limited availability of bio-compatible materials with reliable long-term shape re-tention that are resilient to extrusion or infection hasled to the use of autologous tissue as a preferred sourcefor repair and restoration of craniofacial cartilage. Inthe case of microtia, the gold standard for reconstruc-tion relies on harvesting costal cartilage and sculptingan ear-shaped construct [1]. In rhinoplasty procedures,where a degree of augmentation is required, autologouscartilage tissue can be utilized. Although biocompatibil-ity and rejection are not a concern using autologous tis-sue, harvesting procedures are associated with varyingdegrees of donor site morbidities that can lead to long-term complications. In other settings, limited or lackof tissue availability in post-traumatic cases leave nooption but for the use of FDA-approved biomaterials.

High density porous polyethylene (HDPPE) has beenextensively used for auricle and nasal reconstruction[2, 3]. Although satisfactory results from reconstructiveprocedures using HDPPE can be achieved for nasal andauricle reconstruction, this is an inert inflexible

0022-4804/$36.00� 2010 Elsevier Inc. All rights reserved.

FIG. 1. Porous PVA human shaped auricle. Unlike commerciallyavailable polyethylene implants, the engineered PVA hydrogel ishighly flexible and its porosity allows for seeding of chondrocytesthroughout the construct. (Color version of figure is available online.)

JOURNAL OF SURGICAL RESEARCH: VOL. 163, NO. 2, OCTOBER 2010332

material available in predetermined sizes and shapesand is predisposed to extrusion if implanted undera poorly vascularized environment. Furthermore,implants exposed to traumatic events will invariablyresult in construct extrusion due to the limitedpotential of the surrounding soft tissues to heal overthe implant. The development of a biomimetic andflexible scaffold—particularly important in theauricular and nasal anatomy—could potentiallyimprove the outcomes of surgical procedures anddecrease post-operative complications.

Poly(vinyl alcohol) (PVA) is a biocompatible syntheticpolymer that possesses great potential for the creationof synthetic cartilage [4]. In the surgical field, PVA-based polymers have been used for the prevention ofpostsurgical tissue adhesions and embolization of theuterine artery [5, 6]. Advantages of this materialinclude the versatile formulations that can be created,including that of a porous hydrogel, thus creating anoptimal cell-scaffold environment setting for the fabri-cation of tissue with high water content similar to car-tilage. Furthermore, PVA can be synthesized to beflexible and molded into any size or shape, simulatingthe mechanics of craniofacial cartilage and allowingfor implant customization (Fig. 1).

The goals of this study were to (1) engineer a biosyn-thetic construct using a combination of alginate hydro-gel, a porous non-degradable PVA gel and human nasalseptum chondrocytes, and (2) to quantify and comparethe histologic, biochemical, and biomechanical composi-tions of engineered constructs cultured under differentconditions.

MATERIALS AND METHODS

PVA Hydrogel Preparation

The synthesis of PVA hydrogel has been previously published [7].Briefly, hydrogel was prepared by theta gelation by dissolving PVA(115,000 g/mol) and polyacrylamide-co-acrylic acid (PAAm-co-AAc)and poly(ethylene glycol) (PEG) in deionized (DI) water at 90�C.The resulting solution was then molded into sheets measuring2.1mm in thickness and cooled down to room temperature for gelationfor 24 h. The gel was immersed in DI water for equilibrium and sub-jected to e-beam sterilization.

Cartilage Harvest, Chondrocyte Isolation, and Expansion

All actions were approved by the Institutional Review Board andthe Institutional Animal Care and Use Committee of the Massachu-setts General Hospital and followed all of the policies outlined inthe NIH Guide for the Care and Use of Laboratory Animals. Nasalseptum cartilage from a 34-y-old female patient was collected fromthe operating room. Cartilage was digested and chondrocytes wereisolated as previously described [8]. Briefly, cartilage was washedwith phosphate buffered saline (PBS) twice and minced into 1 mm3

pieces using razor blades. The minced tissue was placed into 50 mLpolypropylene tubes and 40 mL of HAM’S F-12 solution containing0.1% collagenase type II was added. The tubes were placed on a rocker

inside an incubator at 37 �C for 16 h. The contents were then filteredthrough a 100 mm filter and washed twice with PBS. The isolatedchondrocytes were divided into 1.5 3 106 aliquots and placed in 150cm3 cell culture flasks with HAM’S F-12 media with L-glutamine sup-plemented with 10% fetal bovine serum, 50 U/mL penicillin, 50 mg/mLof streptomycin, 50 mg/mL ascorbic acid and 0.1mM nonessentialamino acids mix. Ninety percent confluence was achieved. Aftera 14 d expansion period, the non-passaged cells were enzymaticallydetached using 0.05% trypsin EDTA and used for the construct prep-aration.

PVA Constructs Preparation, Spinner Flask Assembly, and

Construct Implantation

For the construct preparation, chondrocytes were mixed with a 2%sodium alginate (PRONOVA UP LVG) (NovaMatrix, Sandvika, Nor-way) sterile solution and 40 mL of the alginate-chondrocyte mix wasinjected into twelve porous PVA constructs using a 25-gauge needle,resulting in a final cell concentration of 60 3 106 per mL. Gelationwas achieved by immersion into CaCl2 (100 mg/mL) for 5 s. The finalconstruct dimensions (n ¼ 12) were 6 mm diameter and 2.1 mm inthickness. From the 12 constructs, six (group not exposed to a bioreac-tor system) were washed twice in phosphate buffered saline (PBS) andimplanted into the dorsum of nude mice for 6 wk.

The six remaining constructs were placed in a spinner flask bioreac-tor system. For the spinner flask assembly, the spinner flask, needle,silicone stopper, and magnetic stirrer were autoclaved and assembledas previously described [9]. Three constructs were threaded onto each13 cm long stylet of a 22-gauge spinal needle. A total of two stylets, eachwith three constructs, were then secured to the silicone stopper. A 4 cmlong stir bar was inserted into the spinner flask and 120 mL of cell me-dia containing the previously described contents was added. The flaskwas placed on a magnetic stir plate at 50 rpm at 37 �C and 5% CO2 in-cubator. The side caps of the flask were loosened to permit gas ex-change and cell media was changed every third d. After a 10d in vitro period the constructs were implanted into the dorsum ofnude mice. After a 6 wk period in vivo, constructs from both groupswere explanted and grossly examined, subjected to histologic, immu-nohistochemical, biochemical, and biomechanical analysis.

Cell-alginate solution was also used to create nodules (n ¼ 6). Forthe nodule creation, 100 mL of CaCl2 was placed in the bottom of

FIG. 2. Explanted alginate-chondrocyte nodule. Mass grossly re-sembled cartilage tissue after a 6-wk in vivo period. Note the irregu-larly formed shaped due in part by the gelatation kinetics ofalginate hydrogel. (Color version of figure is available online.)

BICHARA ET AL.: HYBRID CONSTRUCT FOR NEOCARTILAGE FORMATION 333

a sterile Eppendorf tube. Then, 400 mL of the cell-alginate solutionwas slowly placed on top using a pipette, and another 100 mL ofCaCl2 was placed on top of the cell-alginate mix, creating a CaCl2-cell/alginate-CaCl2 three-layered structure. After 30 s, the constructswere removed from the CaCl2 layers, washed twice in PBS, and im-planted into the dorsum of nude mice for 6 wk. Upon explantation,the constructs were grossly examined and subjected to histologicaland immunohistochemical analysis.

For the construct implantation, female nude (nu/nu) mice were ob-tained at 6 wk of age and allowed to acclimate for 1 wk. Under asepticconditions, intraperitoneal anesthesia with tribromoethanol (400 mg/kg) was achieved, and a 1.5 cm midline incision was made on the dor-sum of each mouse. Two subcutaneous pockets were created throughblunt dissection using Stevens tenotomy scissors. A total of two con-structs was implanted in each mouse. The incision was closed usingstainless steel Autoclip staples that were removed after 10 d.

Histologic and Immunohistochemical Analysis

Briefly, constructs were fixed in 10% phosphate-buffered formalinovernight, embedded in paraffin, and serial 5 mm sections were ob-tained. Slides were deparaffinized and stained with safranin O and to-luidine blue. For the immunohistochemical analysis, deparaffinizedsections were treated with 2% bovine testicular hyaluronidase atroom temperature for 30 min. A blocking reagent consisting of 0.3%hydrogen peroxide in methanol was added for 30 min. Then, 10%goat serum was added to each slide for 30 min. On separate slides,the primary antibodies for collagen type I (COL I, Abcam, Cambridge,MA, USA) and II (COL II, Chondrex, Inc., Redmond, WA, USA) wereapplied at room temperature for one hour. Both antibodies were di-luted 1:1000 in 1% bovine serum albumin in PBS. For negative con-trol, N-Universal Negative Control was applied. The secondaryantibody (horseradish peroxidase-labeled polymer) was added to theslides for 20 min at room temperature. Then 3,3-diaminobenzidinewas applied to each slide for color development, and cell nuclei werecounterstained with hematoxylin.

Biochemical Evaluations

The quantification of deoxyribonucleic acid (DNA) and glycosami-noglycans was performed using the methodology previously described[10, 11]. The pieces were weighed before and after the 24 hlyophilization period (n ¼ 3 for each group). The dehydratedspecimens were digested by adding 1 mL solution with 100 mMNaPO4, 10 mM Na ethylenediaminetetraacetic acid/disodium salt/dihydrate, 10 mM cysteine hydrochloride, 10 mMethylenediaminetetraacetic acid, and 125 mg/mL of papain. Thespecimens were incubated in a water bath at 60 �C for 16 h. For theglycosaminoglycan quantification, 30 mL of papain digest wereadded to 200 mL of 1,9-dimethylmethylene blue dye and measuredwith a spectrophotometer at 525 nm. The DNA contents werequantified by fluorescence at 360/465 nm after being mixed withbisbenzimidazole dye. The hydroxyproline content of insolublecollagen was determined according to Stegemann and Stalder andnormalized to the dry weight [12]. A portion of papain-digested sam-ple was acid hydrolyzed, neutralized, and reacted with prepare oxidiz-ing solution of chloramines-T and Ehrlich’s solution. All samples andstandards were analyzed in duplicate. DNA, glycosaminoglycan, andhydroxyproline contents were expressed as mg/mg of dry weight.

Compression Strain

Compression strain was performed at the same day of explantation.Mechanical characterization of the biosynthetic constructs was per-formed using an Instron Universal Testing Apparatus (Norwood,MA, USA) (n ¼ 3 for each group). Compressive stress and strainwere calculated and plotted as a stress-strain diagram for bothPVA-alginate-cell construct groups and for the PVA hydrogel alone.

Statistical Analysis

Biochemical quantifications are expressed as mean 6 SD. Sta-tistical significance was determined by ANOVA single factor withP < 0.05.

RESULTS

Gross Examination

Explanted PVA constructs were indistinguishablebetween groups. As expected with a non-degradable hy-drogel, the original dimensions were retained in 12/12explants. There were no signs of extrusion, inflamma-tion, or foreign-body reaction. A fibrous capsule sur-rounded the constructs and was easily removed.Subsequent dissection allowed for construct palpation.Compared with PVA hydrogel alone, described asa sponge-like material, the biosynthetic constructswere viscoelastic when digitally compressed, similarto auricular cartilage. Implanted cell-alginate noduleswere explanted as a solid white glossy irregularlyshaped mass resembling cartilage tissue (Fig. 2). Thesenodules served as a positive control for neocartilage for-mation.

Histologic and Immunohistochemical Analysis

Light microscopy revealed the presence of the non-degradable PVA gel infiltrated with neocartilage tissue.The main structures identified on the PVA gel were hol-low spheres and open channels. The alginate-cell mixwas injected into the channels that are now occupiedby chondrocytes and extracellular matrix. There wasan absence of an inflammatory infiltrate and the tissuewas in close contact with the PVA gel. In all six con-structs that were cultured in the spinner flask prior toimplantation, intense safranin O and toluidine bluestains demonstrated abundant glycosaminoglycan de-position (Fig. 3A and B). Staining for COL II was alsopresent (Fig. 3C). In the group not exposed to

FIG. 3. Histologic and immunohistochemical analysis. Top three images: two structures are visualized including open channels infiltratedwith neocartilage tissue and non-collapsible hollow spheres. Bottom three images: neocartilage formation from the alginate-chondrocyte nod-ules. Stains: A, D ¼ safranin O; B, E ¼ toluidine blue; C, F: collagen type II. (Color version of figure is available online.)

JOURNAL OF SURGICAL RESEARCH: VOL. 163, NO. 2, OCTOBER 2010334

a bioreactor, glycosaminoglycan deposition was alsoidentified by a less intense stain. For the cartilage nod-ules, safranin O and toluidine blue stains revealed gly-cosaminoglycan deposition, identical to the tissuedescribed before that was housed by the PVA hydrogel.Staining for COL II was abundant and revealed a simi-lar intensity compared with the tissue throughout thePVA (Fig. 3D, E, and F).

Biochemical Analysis

All biochemical quantifications normalized to the dryweight of the engineered tissues can be visualized inFig. 4. The biochemical composition of the constructsthat were implanted in vivo immediately after assem-bly was compared with those cultured in a bioreactorprior implantation. The constructs exposed to a bioreac-tor demonstrated increased levels of DNA (5.63 6 0.71mg/mg versus 4.616 0.43 mg/mg), glycosaminoglycans

FIG. 4. DNA, glycosaminoglycan and hydroxyproline quantificationimplanted without bioreactor exposure. Bioreactor þ in vivo ¼ group wias mean þ SEM.

(12.31 6 2.03 mg/mg versus 10.67 6 2.28 mg/mg), and hy-droxyproline (20.67 6 4.5 mg/mg versus 15.49 6 2.3 mg/mg). The increased levels in the group exposed to thebioreactor, expressed as a percent, were the following:22% more DNA, 15% more glycosaminoglycans, and33% more hydroxyproline. Hydroxyproline, an indica-tion of total collagen content in constructs, was foundto be statistically significant (P < 0.05).

Compression Strain

The synthesized PVA hydrogel had a compressiveequilibrium modulus of 25 kPa. Constructs exposed toa bioreactor demonstrated a 22% higher compressivemodulus compared with those that were implantedimmediately after assembly (P< 0.05) (Fig. 5). Additionof the cell-alginate component to the PVA hydrogel sig-nificantly increased the compressive modulus after thein vivo period.

between groups: Control ¼ PVA gel only (acellular). In vivo ¼ groupth a 10 d bioreactor culture period prior implantation. Data presented

FIG. 5. Stress-strain curves. Control¼ porous PVA hydrogel. In vivo¼ group implanted without bioreactor exposure. Bioreactorþ in vivo¼group with a 10 d bioreactor culture period prior implantation. (Color version of figure is available online.)

BICHARA ET AL.: HYBRID CONSTRUCT FOR NEOCARTILAGE FORMATION 335

DISCUSSION

PVA is a versatile material that has been extensivelyresearched for engineering synthetic articular carti-lage. Our group has developed flexible PVA hydrogelsformulations with the potential for replacing defectivecartilage with a synthetic cartilage-like material[7, 13]. However, the subcutaneous use of these inertacellular gels could pose similar complicationsexperienced with HDPPE, including extrusion ifplaced under a poorly vascularized environment, andthe limited capacity of tissues to heal over the implantafter direct trauma. In this preliminary study, wedecided to combine a natural and a novel porous, non-degradable synthetic hydrogel—engineering a biosyn-thetic construct—to evaluate the neocartilage-formingpotential of human chondrocytes derived from the nasalseptum. Furthermore, we evaluated if a 10 d period ina bioreactor would allow ECM deposition and in turnimprove the histologic, immunohistochemical, biochem-ical, and biomechanical properties of the constructs.

These data demonstrate that cartilage can be engi-neered in a subcutaneous environment using a humantissue and a porous PVA-alginate hybrid gel. Due tolower costs and easier access, a vast majority of pub-lished studies regarding engineering of cartilage utilizeyoung healthy animal chondrocytes rather than humantissue. Research reproducibility and utilizing a cellsource that is consistent in both specie and age are ad-vantages of using animal versus human chondrocytes.However, studies using human cell lines are equallyor more important for the field of tissue engineering.In this study, we have used a limited amount of humantissue to engineer cartilage under different conditions.In the constructs exposed to a bioreactor, the shearforce created inside the spinner flask acted againstthe constructs prior in vivo implantation, and increasedthe amounts of DNA and glycosaminoglycans, as wellas the compressive properties. Although subjective,

the intensity of the safranin O, toluidine blue, andCOL II stains correlate with the biochemical and biome-chanical data presented. Unaware of the nutrient diffu-sion properties throughout the PVA gel, the engineeredcartilage nodules served as a positive control for neocar-tilage formation, achieving results similar to nativetissue.

Numerous investigators have researched alginatehydrogel, a biocompatible gel derived from seaweedthat has been previously used to engineer cartilage. Do-bratz et al. injected human nasal chondrocytes sus-pended in 2% alginate gel in a subcutaneousimmunodeficient environment for up to 38 wk [14].The authors wanted to achieve a shape-specific con-struct by using an external mold that would shape thecell-alginate-CaCl2 mix as it was being injected. Al-though neocartilage formation was achieved, the ex-plants were irregularly shaped and did not retaintheir original dimensions over time. Previously, Changet al. also used alginate gel to suspend chondrocytes andengineer auricular cartilage in an ovine model usingautologous cells [15]. In their study, the cell-alginate-CaSO4 mix was injected into a mold shaped likea nose bridge and chin implant. The constructs were im-planted for 30 wk and surprisingly maintained theiroriginal dimensions at the time of explantation. How-ever, it is possible that the alginate gel has not fully de-graded contributing to the retention in size and shape.Described by Kuo and Ma, the gelation kinetics of algi-nate are difficult to control, largely due to the fast cross-linking that occurs when in contact with calciumcations [16]. This results in the inability to form intri-cate shapes that require a defined structure upon im-plantation. For this reason, our group decided tosynthesize a non-degradable porous PVA gel with a pre-defined shape that could subsequently be infiltratedwith alginate gel and chondrocytes.

In tissue engineering approaches where an intricateformed shape is desired, such as a human shaped

JOURNAL OF SURGICAL RESEARCH: VOL. 163, NO. 2, OCTOBER 2010336

auricle, the use of a non-degradable PVA hydrogel pos-sesses numerous advantages over other degradable hy-drogels or polymers. Previously utilized biomaterialsused to engineer human shaped auricles range from fi-brin gel [17, 18] to polylactic acid/polyglycolic acid [19]to poly-l-lactide/poly-3-caprolactone copolymer blends[20, 21]. Although the degradation rate of thesematerials can be controlled and neocartilageengineering can be achieved, shape retention,specifically of an intricately formed shape under anenvironment with overlying tensile forces, is unlikely.The use of a non-degradable material with adequatemechanical properties would be more adequate in thisscenario. Our group has extensive experience using hy-drogels, including collagen [22], fibrin [8], poly(ethyleneglycol) [23], and hyaluronic acid [24]. However, this isour first experience using alginate in combinationwith a non-degradable hydrogel. To our knowledge,this is the first report describing the use of a flexible po-rous biocompatible PVA hydrogel in combination witha natural alginate hydrogel and chondrocytes derivedfrom human nasal septum.

Unaware of the effects of shear force acting againstand throughout the PVA gel, a set of constructs wereplaced in a simple bioreactor system. If future clinicalapplications include placing the hybrid gel under anarea of tension, such as scarred or scarring tissue, im-planting a construct with ECM-producing chondrocytescan potentially be beneficial over time, preventing con-struct deformation from the overlying contractile forcesby the skin. In this study, compression stress testingvaried among groups. Biosynthetic constructs exposedto shear force exhibited a 22% higher (P < 0.05) com-pressive modulus compared with the unexposed groupthat was implanted immediately after assembly.

In conclusion, successful engineering of cartilage hasbeen demonstrated using a novel flexible PVA-alginatehybrid gel scaffold using a limited amount of human tis-sue. The engineered neocartilage histologically resem-bles native tissue, and exposing constructs to shearforce for 10 d in a spinner flask increases the DNA, gly-cosaminoglycans, and hydroxyproline contents, as wellas the mechanical integrity.

ACKNOWLEDGMENTS

This study was supported in part by the American Society of Max-illofacial Surgeons Research Grant and the Department of Orthopae-dic Surgery Academic Enrichment Fund of the MassachusettsGeneral Hospital.

REFERENCES

1. Tollefson TT. Advances in the treatment of microtia. Curr Opin-ion Otolaryngol Head Neck Surg 2006;14:412.

2. Romo T, Reitzen SD. Aesthetic microtia reconstruction withMedpor. Facial Plast Surg FPS 2008;24:120.

3. Peled ZM, Warren AG, Johnston P, et al. The use of alloplasticmaterials in rhinoplasty surgery: A meta-analysis. Plast Recon-struct Surg 2008;121:85e.

4. Oka M, Ushio K, Kumar P, et al. Development of artificial artic-ular cartilage. Proceedings of the Institution of Mechanical En-gineers. Part H. J Eng Med 2000;214:59.

5. Metwally M, Cheong Y, Li TC. A review of techniques for adhe-sion prevention after gynecological surgery. Curr OpinionObstet Gynecol 2008;20:345.

6. Spies JB, Allison S, Flick P, et al. Polyvinyl alcohol particles andtris-acryl gelatin microspheres for uterine artery embolizationfor leiomyomas: Results of a randomized comparative study. JVasc Intervent Radiol 2004;15:793.

7. Bodugoz-Senturk H, Macias CE, Kung JH, et al. Poly (vinyl alco-hol) acrylamide hydrogels as load-bearing cartilage substitute.Biomaterials 2009;30:589.

8. Mesa JM, Zaporojan V, Weinand C, et al. Tissue engineering car-tilage with aged articular chondrocytes in vivo. Plast Recon-struct Surg 2006;118:41. discussion 50.

9. Vunjak-Novakovic G, Obradovic B, Martin I, et al. Dynamic cellseeding of polymer scaffolds for cartilage tissue engineering. Bi-otechnol Prog 1998;14:193.

10. Kim YJ, Sah RL, Doong JY, et al. Fluorometric assay of DNA incartilage explants using Hoechst 33258. Anal Biochem 1988;174:168.

11. Enobakhare BO, Bader DL, Lee DA. Quantification of sulfatedglycosaminoglycans in chondrocyte/alginate cultures, by use of1,9-dimethylmethylene blue. Anal Biochem 1996;243:189.

12. Stegemann H, Stalder K. Determination of hydroxyproline.Clinica Chimica Acta 1967;18:267.

13. Bodugoz-Senturk H, Choi J, Oral E, et al. The effect of polyeth-ylene glycol on the stability of pores in polyvinyl alcohol hydro-gels during annealing. Biomaterials 2008;29:141.

14. Dobratz EJ, Kim SW, Voglewede A, et al. Injectable cartilage:Using alginate and human chondrocytes. Arch Facial PlastSurg 1995;11:40.

15. Chang SC, Tobias G, Roy AK, et al. Tissue engineering of autol-ogous cartilage for craniofacial reconstruction by injection mold-ing. Plast Reconstr Surg 2003;112:793. discussion 800.

16. Kuo CK, Ma PX. Ionically crosslinked alginate hydrogels as scaf-folds for tissue engineering: Part 1. Structure, gelation rate andmechanical properties. Biomaterials 2001;22:511.

17. Ting V, Sims CD, Brecht LE, et al. In vitro prefabrication of hu-man cartilage shapes using fibrin glue and human chondrocytes.Ann Plast Surg 1998;40:413. discussion 420.

18. Xu J, Johnson TS, Motarjem PM, et al. Tissue-engineeredflexible ear-shaped cartilage. Plast Reconstr Surg 2005;115:1633.

19. Liu Y, Zhang L, Zhou G, et al. In vitro engineering of human ear-shaped cartilage assisted with CAD/CAM technology. Biomate-rials 2010;31:2176.

20. Isogai N, Morotomi T, Hayakawa S, et al. Combinedchondrocyte-copolymer implantation with slow release of basicfibroblast growth factor for tissue engineering an auricular car-tilage construct. J Biomed Mat Res Part A 2005;74:408.

21. Kusuhara H, Isogai N, Enjo M, et al. Tissue engineering a modelfor the human ear: Assessment of size, shape, morphology, andgene expression following seeding of different chondrocytes.Wound Repair Regeneration 2008;17:136.

22. Ibusuki S, Papadopoulos A, Ranka M, et al. Engineering carti-lage in a photochemically crosslinked collagen gel. J KneeSurg 2009;22:72.

23. Elisseeff J, Anseth K, Sims D, et al. Transdermal photopolyme-rization of poly(ethylene oxide)-based injectable hydrogels fortissue-engineered cartilage. Plast Reconstr Surg 1999;104:1014.

24. Chung C, Mesa J, Miller GJ, et al. Effects of auricular chondro-cyte expansion on neocartilage formation in photocrosslinkedhyaluronic acid networks. Tissue Eng 2006;12:2665.