Vrije Universiteit Brussel

Indirect rapid prototyping: opening up unprecedented opportunities in scaffold design andapplicationsHouben, Annemie; Van Hoorick, Jasper; Van Erps, Jürgen Albert; Thienpont, Hugo; VanVlierberghe, Sandra; Dubruel, PeterPublished in:Ann. Biomed. Eng.

DOI:10.1007/s10439-016-1610-x

Publication date:2017

Document Version:Final published version

Link to publication

Citation for published version (APA):Houben, A., Van Hoorick, J., Van Erps, J. A., Thienpont, H., Van Vlierberghe, S., & Dubruel, P. (2017). Indirectrapid prototyping: opening up unprecedented opportunities in scaffold design and applications. Ann. Biomed.Eng., 45(1), 58-83. https://doi.org/10.1007/s10439-016-1610-x

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portalTake down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Download date: 10. Nov. 2020

Additive Manufacturing of Biomaterials, Tissues, and Organs

Indirect Rapid Prototyping: Opening Up Unprecedented Opportunities

in Scaffold Design and Applications

ANNEMIE HOUBEN,1 JASPER VAN HOORICK,1,2 JURGEN VAN ERPS,2 HUGO THIENPONT,1,2

SANDRA VAN VLIERBERGHE,1,2 and PETER DUBRUEL1

1Polymer Chemistry & Biomaterials Group, Department of Organic and Macromolecular Chemistry, Ghent University,Krijgslaan 281, S4-Bis, 9000 Ghent, Belgium; and 2Brussels Photonics Team, Department of Applied Physics and Photonics,

Vrije Universiteit Brussel, Pleinlaan 2, 1050 Elsene, Belgium

(Received 8 February 2016; accepted 4 April 2016; published online 14 April 2016)

Associate Editor Jos Malda oversaw the review of this article.

Abstract—Over the past decades, solid freeform fabrication(SFF) has emerged as the main technology for the productionof scaffolds for tissue engineering applications as a result of thearchitectural versatility. However, certain limitations have alsoarisen, primarily associated with the available, rather limitedrange of materials suitable for processing. To overcome theselimitations, several research groups have been exploring novelmethodologies through which a construct, generated via SFF,is applied as a sacrificial mould for production of the finalconstruct. The technique combines the benefits of SFF tech-niques in terms of controlled, patient-specific design with alarge freedom in material selection associated with conven-tional scaffold production techniques. Consequently, well-defined 3D scaffolds can be generated in a straightforwardmanner from previously difficult to print and even ‘‘unprint-able’’materials due to thermomechanical properties that donotmatch the often strict temperature and pressure requirementsfor direct rapid prototyping. These include several biomateri-als, thermally degradable materials, ceramics and composites.Since it can be combined with conventional pore formingtechniques, indirect rapid prototyping (iRP) enables thecreation of a hierarchical porosity in the final scaffold withmicropores inside the struts. Consequently, scaffolds andimplants for applications in both soft and hard tissue regen-eration have been reported. In this review, an overview ofdifferent iRP strategies and materials are presented from thefirst reports of the approach at the turn of the century until now.

Keywords—Indirect 3D printing, Lost-mould, Indirect solid

free form fabrication, Indirect rapid prototyping, Tissue

engineering.

ABBREVIATIONS

2D Two dimensional2PP Two photon polymerization3D Three dimensional3DP Three dimensional printinglm MicrometerAJS Air jet solidificationAM Additive manufacturingBMP Bone-morphogenic proteinsCAD Computer aided designCAM Computer aided manufacturingDDP Drop on demand printingDLP Digital light projectionCPD Critical point dryingCT Computed tomographyDMD Digital micromirror deviceECM Extra cellular matrixeiRP External indirect rapid prototypingFDM Fused deposition modellingHA HydroxyapatiteHFF Human foreskin fibroblastsIJP Ink-jet printingiiRP Internal indirect rapid prototypingiRP Indirect rapid prototypingLoC Lab on chipMEW Melt electrospinning writingmPa Milli PascalMSTL Micro stereolithographyPCL Poly-(e-caprolactone)PDL Periodontal ligamentPDMS Poly(dimethylsiloxane)PEG Poly-(ethylene glycol)

Address correspondence to Peter Dubruel, Polymer Chemistry &

Biomaterials Group, Department of Organic and Macromolecular

Chemistry, Ghent University, Krijgslaan 281, S4-Bis, 9000 Ghent,

Belgium. Electronic mail: peter.dubruel@ugent.beAnnemie Houben and Jasper Van Hoorick have contributed

equally to this work.

Annals of Biomedical Engineering, Vol. 45, No. 1, January 2017 (� 2016) pp. 58–83

DOI: 10.1007/s10439-016-1610-x

0090-6964/17/0100-0058/0 � 2016 Biomedical Engineering Society

58

PGA Poly-(glycolic acid)PLA Poly-(lactic acid)PLGA Poly-(lactic-co-glycolic acid)PPF Poly-(propylene fumarate)RP Rapid prototypingSFF Solid freeform fabricationSLA StereolithographySLS Selective laser sinteringSMD Selective mould dissolutionTCP Tricalcium phosphateUV Ultra violetHUVECs Human umbilical vein endothelial cells

INTRODUCTION

Tissue Engineering

Over the course of history, mankind has always hada desire to prolong life or increase life-quality. Con-sequently, treatment of large tissue defects and organfailure has always been a popular field of study.61

Currently, the major treatment for organ/tissue failureconsists of transplantation or implantation of a pros-thetic device. However, problems typically arise usingthis approach such as scarcity of suitable donor tissueand complications associated to intense immunosup-pressive treatments. Furthermore, prosthetic devicesare often characterized by mismatched mechanicscombined with wear and tear, resulting in mechanicalstress on the surrounding tissue, inflammation andnecessity for repeated surgery. To tackle the increasingamount of these commonly observed problems, tech-nological advances have resulted in alternativeapproaches referred to as ‘tissue engineering’. Tissueengineering has been defined by Langer and Vacanti as‘‘an interdisciplinary field that applies the principlesfrom engineering and life sciences toward the devel-opment of biological substitutes that restore, maintainor improve tissue function.’’39 Of crucial importance inthis field of technology, is the generation of 3D cultureenvironments which mimic the natural extracellularmatrix (ECM) both in terms of mechanics and biologyto grow 3D tissue constructs starting fromcells.24,38,59,67,68,70,107 These so-called scaffolds shouldideally guide cell growth towards the envisionedregenerated tissue in terms of architecture and cellularbehaviour.12,53 Non-toxic, mostly biodegradablebuilding blocks are essential.51 Consequently, healthycells can gradually substitute the scaffold with naturalECM during biosynthesis.19,51 Generally, highlyinterconnected porous scaffolds are considered bene-ficial to facilitate diffusion-driven transport of oxygen,nutrients and waste towards and away from the cellsrespectively, which is only possible over short distances(100–200 lm).3,4,12,71 Additionally, the presence of an

interconnected porous network drastically increasescell seeding efficiency, which results in more homoge-nous cellular distributions throughout the scaffold.Some research groups circumvent hurdles associatedwith inhomogeneous cellular distribution via theincorporation of cells during scaffold produc-tion.1,3,4,23,35,58,62,66,69 This can for example beaccomplished by encapsulating cells in a hydrogelprecursor solution prior to crosslinking To obtain thisbeneficial porosity, several techniques can be appliedincluding conventional techniques (including but notlimited to soft particle leaching, electrospinning,65,107

supercritical CO2 treatment) and more recently solidfreeform fabrication (SFF) techniques.

Solid Freeform Fabrication

SFF is a general term comprising all techniques thatenable the production of structures via the sequentialdelivery of energy and/or materials according to acomputer-aided design (CAD).5 However, often otherterms are applied to describe the same principleincluding additive manufacturing (AM), rapid proto-typing (RP) and 3D printing (3DP).5 RP and AM canbe considered general terms, whereas 3DP only refersto one subcategory based on the main technologicalprinciple.5 These subcategories include irradiation-based systems, nozzle-based systems and printer-basedsystems.

Currently, SFF is considered as the ‘‘golden stan-dard’’ for the generation of scaffolds due to its signif-icant benefits over conventional porous scaffoldproduction technologies. In conventional approachesthe micro-architecture is mainly process-determinedrather than design-driven and often requires extensivemanual manipulation.12,56 Furthermore, the obtainedporosity is often random and poorly interconnected,resulting in limited transport of oxygen and nutri-ents.71,77 In addition, industrially applied fabricationmethods such as injection moulding or hot embossing,can never generate equally complex internal struc-tures.12 In contrast, SFF enables precise control overboth macrostructure and microstructure with a highdegree of reproducibility and homogeneity.12,13,56,89

Additionally, the architecture of the micropores caninfluence cellular behaviour.74,89 As a result, SFF en-ables superior control over diffusion-driven transport,cell growth and differentiation.51,53,74 Hence, themechanical properties of the scaffolds can be tailoredby design. Even gradient mechanical properties can begenerated by combining different pore sizes/architec-tures within one single construct.27,30 Moreover, theunique straightforward translation from CAD toCAM allows to tailor each implant to meet specificneeds/designs often based on a 3D model obtained

Indirect Rapid Prototyping 59

from medical imaging techniques.40,84,93,104 Conse-quently, since the construction of each specific implantdoes not require the expensive production of robustmolds, patient specific implants become cost-efficientenough for translation to the clinic.16,84In conclusion,SFF is the preferential technique for the generation ofcost-efficient, patient-specific constructs closely mim-icking the tissue defect based on medical imagingtechniques.12,82,93,104

However, SFF techniques also exhibit certain limi-tations. For instance, the application of laser-basedsystems is limited to photo-polymerizable materials(e.g., Stereolithography, SLA or Two photon poly-merization, 2PP) or highly controllable melting mate-rials (Selective Laser Sintering, SLS). In addition,unreacted monomers and photo-initiators applied inSLA can exhibit cytotoxicity.12 Constructs generatedvia printer based systems are often very brittle andmostly require a post-processing treatment to increaserobustness.5,37,54 As for nozzle-based systems, certaindesign limitations exist. For example, the resolution isinferior to printer based systems and irradiation-basedtechniques. Also, certain shortcomings can be attrib-uted to SFF techniques in general. First, each SFFtechnique is only suitable for the processing of a lim-ited range of materials which exhibit the requiredproperties. Secondly, resolution and architecturalcontrol are often highly dependent on applied materialand SFF technique. Finally, successful SFF of novelmaterials frequently requires extensive and time-con-suming optimization of processing parameters.5,63,87

Therefore, compromises often have to be made interms of resolution or material properties to generatesuitable constructs.4

Indirect Rapid Prototyping

To overcome these limitations, several researchgroups have been and are exploring a methodologywhich combines the benefits of SFF in terms of con-trolled, patient-specific design with the freedom ofmaterials associated with conventional scaffold pro-duction techniques.20,30,34,36 This approach applies aconstruct, generated via SFF techniques as a sacrificialmould or template for the final construct.95 Therefore,the approach is referred to as ‘‘lost-mould technique’’,indirect solid freeform fabrication, indirect rapid pro-totyping (iRP), indirect additive manufacturing orindirect 3D printing.13,29,43,81,82 Here, the processingparameter requirements of the final scaffold materialare of secondary importance, as a mould is generatedusing straightforward processing of well-establishedmaterials.12,34 To generate the template, SFF tech-niques can be applied with superior resolution com-

pared to direct SFF of the desired material.20 By usinghigh resolution techniques for template fabrication, forexample SLA, scaffolds with pore and strut sizes aslow as 50 and 65 lm respectively can be obtained formaterials unsuitable for direct fabrication using highresolution manufacturing methods.34 Consequently,well-defined 3D scaffolds can be generated in astraightforward manner from materials previouslyidentified as being ‘‘unprintable’’ due to thermome-chanical properties that do not match the often stricttemperature and pressure requirements for direct rapidprototyping.10,14,19,40,44,50,81,87 Furthermore, less rawmaterial is required since material consuming processoptimization becomes superfluous. Additionally, onlythe volume of the final construct is required for scaf-fold production whereas most direct SFF techniquesrequire excessive material amounts. For example, inSLA, uncrosslinked material is washed away aftershaping, in 3DP and SLS unbound particles are blownaway after structuring.87 Additionally, iRP allows astraightforward combination of different materialswithin one scaffold, including bioactive compounds.This is achieved either via blending of materials, or viathe creation of zones with different mechanical prop-erties for specific tissue function (e.g., interfacebetween soft and hard tissue).44,87 Moreover, iRP al-lows the creation of a hierarchical porosity in the finalscaffold, since porous microstructures can be obtainedinside the struts via conventionally applied techniquesduring or after casting.10,29,34,36,42–44,50,53,77,79,87

Consequently, iRP has proven to circumvent hur-dles commonly associated with direct SFF techniquesand opens up unprecedented opportunities towardsnew scaffolding strategies 87 (see also ‘‘Generation ofMicropores Inside the Scaffold’’ and ‘‘Applications ofIndirect Rapid Prototyping’’ sections). The combina-tion of a large selection of RP techniques with a vastvariety of casting/template removal strategies results ina large range of diverse iRP strategies. However, thegenerally applied methodology consists of three mainsteps 19,51,53: (Fig. 1).

1. Mould fabrication through SFF.2. Casting of the desired material + fixation of

the final shape3. Mould removal.

The design of the mould has to correspond to theinverse of the final construct.34,40 Often this isaccomplished in silico by generating an inverse 3Dmodel.30 In some cases SFF techniques already apply asecondary material during printing to act as a supportfor the final design, which can be selectively removedafterwards. Hence, selective removal of the constructmaterial rather than support material and application

HOUBEN et al.60

of the support structure as mould for iRP has beenreported.72 Consequently, the process of designing aninverse mould becomes redundant.72,73 Anotherapproach to avoid designing an inverse mould is togenerate a first structure via SFF exhibiting the designof the final scaffold. This structure is then used as asacrificial mould to generate the negative sacrificialmould for the final construct.87–89,93

To classify the range of approaches, a first catego-rization can be obtained by distinguishing between twomain iRP strategies. First, the combination of a sac-rificial mould generated by SFF to control the macro-architecture (i.e., the external shape) of the final con-struct with conventional pore forming has beenreported.42,43,50,53,85,93,104 In the present paper, thesemethodologies will be referred to as external indirectrapid prototyping (eiRP).11 (Fig. 1).

Secondly, the application of SFF templates has beentargeted to obtain a controlled microporous structureinto the final scaffold.14,26,29,30,41,81,82,87,89,95 Thismethodology will be referred to as internal indirectrapid prototyping (iiRP) (Fig. 1). Sometimes conven-tional techniques are applied in combination with iiRPto generate a hierarchical microporosity, resulting inthe generation of porous struts which follow themorphology of the template.41,56,87 Often also a com-bination of both approaches has been applied throughwhich a desired outer contour is merged with well-defined inner architectures.10,30,34,40

INDIRECT RAPID PROTOTYPING: APPLIED

METHODOLOGIES AND MATERIALS

Mould/Template Fabrication Process

Most SFF techniques share a similar process flow.First, a 3D CAD design is generated which can beobtained from medical imaging or designed to meetspecific needs. Next, it is sliced along the Z-axis intoseveral consecutive ‘2D’ layers. Finally, these consec-utive layers are transferred to coordinates and fed tothe SFF device, which processes them into a 3D con-struct in a layer-by-layer fashion. Some SFF tech-niques have the possibility to operate outside of theconventional layer-by-layer convention. This is forexample the case with 2PP where a focussed spot canbe scanned through a photo-crosslinkable solution in 3dimensions. However, these also often apply layer-by-layer fabrication to decrease production times andsimplify CAD processing.92 An overview of currentlyapplied template fabrication processes for iRP is pre-sented in Fig. 2 and the paragraphs below.

Irradiation-Based Systems

Irradiation-based systems deliver energy via irradi-ation to transfer a CAD to a certain material. A dis-tinction can be made based on the purpose of theapplied irradiation between local photo-polymeriza-tion systems (e.g., SLA, 2PP) and sintering systems

FIGURE 1. Comparison of the eiRP and iiRP approach.

Indirect Rapid Prototyping 61

which apply heat to locally fuse particles together(i.eSLS). To date, only photo-polymerization tech-niques have been applied for iRP. In this respect, thebest known technique is SLA where a laser beam isscanned across the surface of a photo-curable resin or(pre)polymer solution, thereby locally polymerizing thematerial. After solidification of a complete layer, theplatform moves downwards for processing of the nextlayer. After multiple iterations, unpolymerized mate-rial is removed leaving behind the polymerized tem-plate.30,59,83

A similar approach is digital light projection (DLP)or microstereolithography (MSTL), where instead oflaser irradiation, a UV lamp is used combined with amicromirror array. Every layer is converted into apixelated image and transferred to the micromirrorarray (digital micromirror device, DMD) where everypixel corresponds to one micromirror.49 Mirrors cor-responding to areas which need to solidify reflect theUV light to the polymer solution, whereas theremaining mirrors reflect the light away.36,106 Conse-quently, each layer is generated in one flash in contrastto processes exploiting the scanning of a laser beam,which makes the technique faster than conventionalSLA.36 (Fig. 3).

High resolution and precise architectural controlrender SLA and DLP/MSTL ideal for complex poroustemplates suitable for iiRP with high shapefidelity.9,14,30,106 Currently, MSTL techniques enablegeneration of structures with (sub)-micron sized fea-tures.84 However, since the applied mould is photo-crosslinked, template removal becomes less straight-forward. Consequently, the method is generally used to

generate ceramic scaffolds which allow template re-moval via pyrolysis.83,106

Soft lithography is another irradiation-based tech-nology utilized in iRP. First, a pattern is generated vialithography on a robust photoresist. Next,poly(dimethylsiloxane) (PDMS) is ‘‘cast’’ onto thegenerated pattern, and the inverse pattern is trans-ferred to the PDMS by hardening to generate a PDMS‘‘stamp’’.23 This soft ‘‘stamp’’ is then coated with thedesired material followed by transfer of this material toa targeted surface, similar to conventional stampscoated with ink. The technique is referred to as softlithography since lithography is applied as a master fora soft stamp material (PDMS). The technique cannotbe considered a true SFF technique since only ‘2.5D’single-layer patterns are generated through thisapproach. However, by introducing several single-layer templates into the final material followed bytemplate removal, 3D scaffolds can be obtained con-taining a stack of planar channels for iiRP.23

Nozzle-Based Systems

Nozzle-based systems apply controlled X–Y move-ment of a material-depositing nozzle to generate eachlayer of the CAD. Increasing distance between theplotting head and plotting platform enables stacking ofthese layers to generate a 3D structure. A subcatego-rization can be performed based on the appliedextrusion/dispensing mechanism.

The most straightforward nozzle-based technique isfused deposition modelling (FDM). The method uses athermoplastic polymer filament which is fed by rollers

FIGURE 2. Overview and classification of currently applied template fabrication techniques for indirect Rapid Prototyping.

HOUBEN et al.62

through a heated nozzle which melts the polymermaterial just before deposition.5,38,55,107 During depo-sition, the heat of the freshly deposited polymer softensthe previous layer thereby inducing interlayer attach-ment.109 FDM can generate structures with strutdiameters in the range of several 100 micrometerswhich is mainly controlled by nozzle diameter andextrusion rate.109 The range of suitable materials forFDM remains limited since the materials have to bethermoplastic, exhibit suitable viscoelasticity andmelting/solidification properties and have to be pre-processed into polymer filaments. Several researchgroups have utilized this technology for the generationof controlled microstructures in iiRP.6,7,27,95

A variation to FDM is air pressure jet solidification(AJS) which applies air pressure instead of rollers toforce solutions or melts through a nozzle, therebyremoving the need for preformed thermoplastic poly-mer filaments. In contrast with FDM, a larger range ofmaterials can be applied including hydrogels, polymer/ceramic sludges, etc.10,20,21 Similar to FDM, strutdiameters can be obtained in the range of several 100micrometers.10

Recently, the use of direct-write assembly wasreported as a high-resolution variant of AJS. Via thecombination of fine nozzle tips with fugitive inks,exhibiting specific viscoelastic properties, fine struts(down to 10 lm diameter) can be obtained by opti-mizing the pressure.90,91 These properties have to en-able ink deposition under high shear conditionscombined with sufficient shape retention after struc-

turing.90,91 Therefore, sometimes the ink is photo-crosslinked during deposition to ensure sufficientmechanical integrity.2 However, since only a limitedrange of materials comply with these stringent vis-coelastic properties, the technique is more suitable foriiRP than direct SFF.90,91,108 More specifically, theability to generate struts with dynamic diameter con-trol by pressure variation during processing makes thetechnique suitable for generating hierarchical vascularnetworks via iRP.108

Other nozzle-based techniques apply a piston for thedeposition of material. In one specific iRP example, aliquid hydrogel precursor is sucked into a capillary usinga piston.3 Next, physical gelation is induced into thecapillary, followed by the deposition of the gel in acontrolled area to obtain linear channels into amatrix ofthe final scaffold material with a diameter of 500 lm.3

Recently, a variation to nozzle based SFF tech-niques referred to as melt electrospinning writing(MEW) has been reported which applies the principlesof melt electrospinning in combination with fluiddynamics to obtain ultra-narrow strut diameters typi-cally in the range of 5–50 lm.8,26 Although the tech-nique is very recent, it has already been applied togenerate microchannels in a hydrogel structure viaiiRP.26

Printer-Based Systems

3DP applies selective deposition of a liquid binderon a powder bed to selectively fuse powder particles

FIGURE 3. Digital light projection/microstereolithography principle.

Indirect Rapid Prototyping 63

together into a single layer of the CAD.40,42,43,54 Aftercompletion of a layer, a roller distributes a new layer ofpowder onto the previous one and the process is re-peated. Since infused powder acts as a support, com-plex 3D structures can be obtained with resolutionsdepending both on jet droplet and powder particle size.For iRP purposes, resolutions in the range of 100–200 lm have been reported.40,42,60 When the completestructure is printed, unfused particles are blown awayusing pressurised air.54 Often, the constructs are verybrittle and require infiltration with a polymer solution(e.g., polyethylene glycol) or PEG) to reduce brittle-ness.54 The material selection is limited, especially theuse of synthetic degradable materials poses problemssince organic solvents required for particle fusion dis-solve most commercially available print nozzles.40,42

This in combination with the poor mechanical prop-erties make the technique more attractive for iRP thandirect SFF if template removal is possible.40,42 There-fore, Lee et al. used gelatin microparticles as a powderfused together with an aqueous solution to generatewater-soluble templates.41

In addition to 3D printing, other printerbasedtechniques apply an inkjet nozzle to directly depositdroplets of a build-material (e.g., wax) by usingpiezoelectric actuation. This technique is referred to asinkjet printing (IJP), drop-on-demand printing (DDP)or phase change 3D jet-printing.51,63,79 Upon impact,the droplets harden thereby forming beads which resultin lines when adjacent beads overlap.51,79 The methodbenefits from a high resolution as a consequence of theinkjet principle.20 Often, two nozzles are applied todeposit two materials, being a build material to repli-cate the CAD and a support material to supportoverhanging struts during processing.44 Application ofa support material requires an additional support re-moval step by using a selective solvent before thetemplate structure is obtained.20,44,50,81 Additionally,some systems increase precision by milling away thetop of each layer to a well-defined height before thenext layer is deposited.47,51,56,81,82,105 (Fig. 4) Usingthis methodology, a layer height in the range of10–20 lm can be obtained.20 Since the applied materialin this technology is mainly limited to specific waxeswhich can easily be dissolved, the technique hasproven to be very popular for iRP as it combi-nes high resolution with straightforward templateremoval.15,21,81,105

A small variation to IJP can be considered as ahybrid between printer- and irradiation-based tech-niques. The technique is referred to as POLYJETTM

and applies deposition of photocurable droplets com-bined with immediate UV-curing, in this manner morerobust structures are obtained in comparison to con-ventional IJP. Similar to IJP, the application of 2

nozzles enables the use of a support structure whichcan be selectively dissolved afterwards.88,89

Casting Methods

Following template generation, a casting step isperformed to introduce the final scaffolding material.Since filling up the entire voids with the final scaf-folding material can be difficult, several strategies haveemerged in this respect. Applied methods includevacuum treatment, injection moulding and thermo-forming.

Vacuum can be applied to remove all air bubblesfrom a solution during scaffold infiltration.44 Onestrategy incorporates several templates into a solutionof the final material followed by placing the entirecontainer under vacuum.95 In another approach asyringe filter has been cut in half to which the templatewas secured. Next, the filter and template were placedinto a solution containing the final scaffold materialfollowed by generating a vacuum using the plunger ofthe syringe thereby resulting in suction of the solutioninto the entire template (Fig. 5).105

If the viscoelastic properties of a melt allowstraightforward material handling, pure polymer meltsare directly injected into the mould via a syringe.54,63

To this end, an extrusion chamber can be incorporatedin the mould design. The chamber can be filled with amelt, followed by reproducible extrusion using a syr-inge plunjer (Fig. 6).63 After template removal, theremnants of the extrusion chamber have to be removedmanually.63

Some research groups have already applied a designfor the sacrificial mould which incorporates an inletwhere a syringe can be attached and an outlet toremove excess material and air (Fig. 7). As a result,the final scaffold material can simply be injected intothe mould, similar to injection moulding.34,36 Conse-quently, scaffold generation is enabled in low techenvironments (e.g., hospitals when it comes to scaf-fold production), if the templates are generated else-where.34

As a final method, Sodian et al. have applied atechnique where porous sheets of the scaffold materialwere heated, pressed against the negative mould andcooled to induce crystallization, thereby locking thefinal shape.85,86 As a consequence, complete heartvalves could be obtained from a single sheet with acontrolled porosity without the need for suturing.86

Generation of Micropores Inside the Scaffold

In order to generate (additional) porosity in thescaffolds produced via iRP, several strategies have beenconsidered to date: critical point drying, freeze drying,

HOUBEN et al.64

porogen leaching, phase inversion, foaming andintroducing gas forming agents.

Critical point drying implies the exchange of anorganic solvent with liquid CO2 at room temperatureat a controlled pressure. Next, the temperature isincreased to 10� above the critical point of CO2 whichconverts liquid CO2 into gas, leaving behind a porousnetwork. The use of an organic solvent, often ethanol,is of crucial importance since liquid CO2 is not misciblewith water. The obtained pores are generally about50 lm large.29

Freeze drying uses water (or another solvent) whichis frozen to obtain ice crystals, followed by sublimationof these crystals. Consequently, empty pores are leftwith the morphological features of the ice crys-tals.29,56,97,98 When the solvent is water, the mean porediameter is typically in the range of 100–200 lm.56

In another approach, soluble salts or porogens canbe incorporated into the scaffold material prior tocasting. During template removal, the porogens areremoved simultaneously, thereby leaving behind poreswhich exhibit the morphology and the size of theapplied porogen.34,36,40–43,85,89 Tan et al. have applieduniform microspheres as porogens and arranged themin the most closely packed conformation via ultrasonictreatment. Using this method, a reproducible porousinterconnective network can be obtained.89

If a homogenous solutionof twoormore components(e.g., PCL inN–N-dimethylacetamide) exists at a certaintemperature, sometimes phase separation between thecomponents can occur during cooling. After precipita-

tion, the solvent can be washed away, leaving behind aporous structure in the precipitated compound (e.g.,poly-e-(caprolactone) or PCL).11,34,36,85 Under the right

FIGURE 4. Principle of phase change 3D printing also referred to as drop on demand- or inkjet printing.

FIGURE 5. Casting via syringe-induced vacuum treatment.Reprinted from Ref. 105 with permission from Wiley.

Indirect Rapid Prototyping 65

conditions (solvent system, cooling conditions), evennanofibrous structures can be generated which mimicthe morphology of collagen fibers present in theECM.11,104

Another technique which has been explored isfoaming. By vigorously stirring certain scaffold mate-rials including gelatin, viscous foams can be obtained.When the scaffold material is cast into the mould fol-lowed by immediate freezing to lock the shape, a highlymicroporous material is obtained with pores revealinga spherical morphology and uniform distribution. Thepore size can be controlled by varying the stirringspeed during foam generation.88

A final reported approach to obtain micropores in aconstruct consists of the application of gas producingcompounds. For example, when radical (photo-)

polymerization is applied to lock the template shapeinto the final material, the use of azo-type initiators(e.g., VA-086) results in the generation of N2 gasduring activation. This results in the formation ofmicropores inside the final construct.4

Mould/Template Removal Process

The final step in iRP consists of template removalwithout damaging the final construct.34 The absence ofchemical interactions between template and scaffoldingmaterial facilitates this process. In addition, elasticityand swelling/shrinking behaviour of both the templateand material should be taken into account. For in-stance, if brittle scaffolding materials are applied,swelling of the template during dissolution can result in

FIGURE 6. Principle of casting via injection moulding. In this approach the template contains a void for the final material beforedirect syringe injection.

FIGURE 7. iRP principle including direct syringe injection casting step. The template contains an inlet for the syringe and anoutlet for removal of excess material.

HOUBEN et al.66

stress-induced cracks and deformations arising in thefinal construct.49 As a result, several techniques havearisen, including selective melting followed by evapo-ration, binder burnout and sintering, selective moulddissolution, selective mould degradation, and manualremoval or aspiration.

If the template material exhibits a melting pointbelow the thermal stability threshold of the scaffoldmaterial (e.g., wax and poly-(glycolic acid)/poly-(lacticacid) or PGA/PLA), the template can be removed byselectively melting the template.87 To ensure completetemplate removal, a solvent can be applied to washaway any residue.87 Sometimes melting alone is notsufficient for complete scaffold removal and oftenevaporation is necessary. In direct write assembly, theapplied template material is often a fugitive ink. As aresult, the template (vascular) structure can beremoved by gentle heating to evaporate the ink.90,91

When ceramics are applied as a scaffolding material,organic templates can be removed by burnout orpyrolysis prior to sintering.6,7,13,14,20,21,30,81,82,105

Pyrolysis followed by evaporation occurs when anorganic material is heated to high temperatures(>400 �C).13,105 As a result, even unsolvable cross-linked networks generated by irradiation-based tech-niques can be removed.9,13,106 As sintering at hightemperatures is essential to remove the binder materialapplied for the ceramic slurry, template removal canoccur without an additional step.36

Selective mould dissolution (SMD) is predomi-nantly applied if the final scaffold material is a (syn-thetic) polymer. A prerequisite of the technique is theavailability of a solvent which selectively dissolves thetemplate material without influencing the scaffoldmaterial.11,29,42–44,50,54,56,63,77,79,87–89,93 Therefore, thetechnique can be applied if the scaffold consists ofcrosslinked polymer networks which are unsoluble.95

In addition, solvent mismatching systems can be anoption (water vs. organic solvents).10,40 The method isoften applied to remove wax templates generated viaDDP/IJP techniques.15,47 Also, the removal of ceramictemplates for the generation of scaffolds by usingceramic-specific solvents (e.g., RDO-APEX Engineer-ing Products Corp) has been reported.87

SMD is one of the few template removal techniquessuitable for cell encapsulation in iRP. This can beaccomplished when a water-soluble template (e.g., ge-latin) is combined with a crosslinked hydrogel materialcontaining cells (e.g., collagen).23

If the combination of high-resolution SLA tem-plates with thermally unstable scaffolding materials(e.g., (bio)polymers) is desired, pyrolysis and SMD arenot possible for template removal.49 Therefore, thesolution can sometimes be found in creative chemistry.Liska et al. reported on the use of a hydrolysable

crosslinkable material (using methacrylic anhydride asa crosslinker) for SLA template generation. Aftercasting, the mould was selectively removed viahydrolysis of the anhydride bonds using an alkalinesolution.49 Using this approach, high-resolution iiRPand eiRP scaffolds could be obtained from (PCL),poly(lactic-co-glycolic acid) (PLGA), poly-(L-lacticacid) (PLLA), chitosan, alginate and bone cement.34

Aspiration is a technique through which the tem-plate is removed from the final scaffolding materialusing a low vacuum pressure. It is clear that thismethod can only be applied for the removal of tem-plate fibers which do not show adhesion to the finalscaffolding material.3 The method can for instance beapplied for the generation of blood vessels in labs on achip (LoC) and organs on a chip.3 Consequently, thetechnique is restricted to the generation of simple un-branched channels, which limit tissue engineeringapplications. However, these properties are often suf-ficient for LoC devices. The method can be consideredas the least invasive template removal strategy.

SMD and aspiration have certain benefits overother scaffold removal methods, since they both allowthe application of cell encapsulation via iRP.3,23,62

Furthermore, the sacrificial scaffolds can be producedelsewhere as a ready-to-go kit. Consequently, cellu-larized scaffolds can be obtained without the necessityof SFF knowhow and infrastructure. This featureincreases the potential for rapid technology adoptionand swift transfer to the clinic.62 Consequently, iiRPposes substantial benefits for the generation of vascu-lar networks in comparison to direct SFF techniques.62

Overview of Scaffold Materials

A range of indirect scaffold materials have beendescribed in literature to date including biopolymers,synthetic polymers, ceramics and composites. Manyauthors have highlighted the versatile nature of theiRP technique by casting a selection of (bio)polymersand ceramics.3,31,62 In addition to (bio)polymers andceramics, the iRP technique can also be used for theconstruction of porous metal implants, enabling cel-lular ingrowth.60 Table 1 gives an overview of all iRPapproaches applied to date and contains informationabout the applied materials, the SFF technique, tem-plate material removal and the secondary pore creationstrategy.

Biopolymers in Indirect Rapid Prototyping

Extracellular matrix proteins are especially inter-esting for tissue engineering purposes, as they areinherently biocompatible, cell-interactive and do notresult in toxic degradation products. However, they are

Indirect Rapid Prototyping 67

TA

BL

E1.

Overv

iew

of

the

dif

fere

nt

mate

rials

an

dm

eth

od

s,

arr

an

ged

by

scaff

old

mate

rial.

Mate

rial

SF

Fte

chniq

ue

Tem

pla

tem

ate

rial—

rem

oval

Secondary

pore

cre

ation

Image

a

Bio

poly

mers

Colla

gen

Soft

Lithogra

phy2

3

IJP

51,7

8,7

9,1

10

Gela

tin—

Melti

ng

and

flushin

g23

Wax-D

issolu

tion

51,7

8,7

9,1

10

Therm

opla

st-

Dis

solu

tion

110

Criticalpoin

tdry

ing

Lyophili

sation

110

[50]

Gela

tin

Bio

printing

3

3D

P89

FD

M95

Agaro

se—

Aspiration

3

Photo

poly

mer-

Dis

solu

tion

89

PLLA

-Dis

solu

tion

95

Low

densi

tyP

Sm

icro

sphere

s89

[89]

Agaro

se

FD

M62

Carb

ohydra

tegla

ss—

Dis

solu

tion

62

Alg

inate

FD

M62

DLP

34

Carb

ohydra

tegla

ss—

Dis

solu

tion

62

Photo

reactive

mix

ture

-Hydro

lysis

34

Fib

rin

FD

M62

Soft

Lithogra

phy2

3C

arb

ohydra

tegla

ss—

Dis

solu

tion

62

Gela

tin-M

eltin

gand

Asp

iration

23

[23]

Matr

igel

FD

M62

Carb

ohydra

tegla

ss—

Dis

solu

tion

62

Chitosan

DLP

34

IJP

44,5

6P

hoto

reactive

mix

ture

-Hydro

lysis

34

Wax-D

issolu

tion

44,5

6C

riticalpoin

tdry

ing

56

Lyophili

sation

44

Fib

roin

IJP

50

Therm

opla

st-

Dis

solu

tion

50

Lyophili

sation

50

HOUBEN et al.68

TA

BL

E1.

co

nti

nu

ed

Mate

rial

SF

Fte

chniq

ue

Tem

pla

tem

ate

rial—

rem

oval

Secondary

pore

cre

ation

Image

a

Synth

etic

poly

mers

PE

GB

ioprintin

g3

IJP

62

Agaro

se—

Aspiration

3

Carb

ohydra

teG

lass

–D

issolu

tion

62

Epoxi

de

Direct

Writin

g91,1

08

Org

anic

ink—

Meltin

gand

Aspiratio

n91,1

08

PC

L3D

P41,5

4

DLP

34,3

6

SLA

31

IJP

63,7

3,8

5,8

6,9

3

Gela

tin—

Dis

solu

tion

41

Photo

reactiv

em

ixtu

re-H

ydro

lysis

34,3

6

Epoxy

-n.a

.31

Pla

ste

r-D

issolu

tion

54

Wax-

Dis

solu

tion

63,7

3

Therm

opla

st-

Sin

tering-H

A(2

nd)-

Dis

solu

tion

85,8

6

Wax-

Dis

solu

tion-P

VA

(2nd)-

Dis

solu

tion

93

Poro

gen

Leach

ing,

Phase

invers

ion

36

Poro

gen

Leach

ing

73,8

5,8

6,9

3

[54]

PLLA

DLP

34

IJP

11,8

7,1

04

SLA

31

Photo

reactive

mix

ture

-Hydro

lysis

34

Wax-

Dis

solu

tion

11,1

04

Epoxy

-n.a

.31

Wax-

Sin

tere

d-H

A(2

nd)-

Dis

solu

tion

87

Lyophili

sation

94

Poro

gen

leachin

g,

em

uls

ion,

freeze

dry

ing

87

Phase

separa

tion

11

Phase

separa

tion,

Poro

gen

leachin

g104

[11]

PG

AS

LA

31

Epoxy-n

.a.3

1

PLG

AD

LP

34

SLA

31

3D

P42,4

3

IJP

64,8

0

Photo

reactive

mix

ture

-Hydro

lysis

34

Epoxy

-n.a

.31

Pla

ste

r-D

issolu

tion

42,4

3

Wax-

Dis

solu

tion

64

Wax-

Sin

tering-H

A(2

nd)-

Dis

solu

tion

80

Poro

gen

leachin

g42,4

3

[80]

PLA

/PG

AIJ

P87

Wax-S

inte

red-H

A(2

nd)-

Dis

solu

tion

87

Poro

gen

leachin

g,

em

uls

ion,

freeze

dry

ing

87

Indirect Rapid Prototyping 69

TA

BL

E1.

co

nti

nu

ed

Mate

rial

SF

Fte

chniq

ue

Tem

pla

tem

ate

rial—

rem

oval

Secondary

pore

cre

ation

Image

a

PG

A/P

CL

IJP

72

Wax-D

issolu

tion

72

pO

xM

EW

26

PC

L26

PP

FIJ

P45

Wax-D

issolu

tion

45

PH

OH

/P4H

BS

LA

85,8

6P

lastic-M

anual8

5,8

6P

oro

gen

leachin

g85,8

6

Cera

mic

s

Bone

cem

ent

DLP

34

Photo

reactive

mix

ture

-Hydro

lysis

34

HA

IJP

20,2

1,3

2,5

6,8

1,8

2,1

05

SLA

13,1

4,3

0,3

1

DLP

74,7

5,8

3,1

06

Wax-D

issolu

tion

34,5

6

Epoxy-S

inte

ring

13,1

4,3

0

Wax-S

inte

ring

20,2

1,3

2,1

05

Epoxy-n

.a.3

1

Photo

reactive

resin

-Sin

tering

74,7

5,8

1–83,1

06

Criticalpoin

tdry

ing

34,5

6

[14]

Alu

min

aF

DM

6,7

Therm

opla

st-

Sin

tering

6,7

TC

PIJ

P15,4

7,8

1,8

2

SLA

31

Wax

-D

issolu

tion

15,4

7

Epoxy-n

.a.3

1

Wax-S

inte

ring

81,8

2

[106]

HA

/TC

Pble

nds

SLA

31

IJP

81,8

2E

poxy-n

.a.3

1

Wax-S

inte

ring

81,8

2

CoC

rMo

n.a

.81,8

2C

era

mic

-n.a

.81,8

2

Com

posites

and

ble

nds

Chitosan/H

A3D

P41

Gela

tin-D

issolu

tion

41

Gela

tin/P

LLA

3D

88

Photo

poly

mer—

Dis

solu

tion

88

Foam

ing

88

HA

/Sucro

se

AJS

10

n.a

.-M

anualre

moval

10

TC

P/P

PF

/NV

PIJ

P/D

DP

15

Dis

solu

tion

15

PLA

/HA

/PG

AIJ

P87

Wax-S

inte

red-H

A(2

nd)-

Dis

solu

tion

87

Poro

gen

leachin

g,

em

uls

ion,

freeze

dry

ing

87

PC

L/T

CP

3D

P54

Pla

ste

r-D

issolu

tion

54

PC

L/C

aP

3D

P63

Wax-D

issolu

tion

63

PP

F/T

CP

SLA

31

IJP

48

Epoxy-n

.a.3

1

Wax-D

issolu

tion

48

aA

llexam

ple

figure

sare

reprinte

dw

ith

perm

issio

nfr

om

the

respect

ive

publis

hers

as

indic

ate

din

the

sourc

e.

HOUBEN et al.70

often considered hard to process. Of considerableinterest for tissue engineering purposes is collagensince it is the major constituent of the natural humanextracellular matrix.23,29,78,79 Fabrication of microchannels in a collagen matrix using conventional RPtechniques has been limited to simple geometries, suchas cylindrical tubes. To generate collagen scaffoldswith complex microfluidic channels, a gelatin moulddeveloped by stacking several patterned layersobtained via soft lithography has already beenapplied.23 For some applications, the presence ofmicropores inside the scaffold can pose benefits withrespect to diffusion-driven transport. When comparingcritical point drying (CPD) and freeze drying for col-lagen scaffolds, CPD led to significantly highershrinkage of the scaffold and could not result in afoam-like porous structure. Therefore, freeze drying isconsidered a more suitable technique for introducingmicroporosity in collagen scaffolds.29

Gelatin is a material derived from collagen either byacidic or alkaline hydrolysis. It has been combinedwith microspheres in an iRP approach to createchannelled, microporous scaffolds.89 However, gelatindissolves rapidly at physiological conditions and doesnot possess sufficiently high mechanical propertiesneeded for many tissue engineering applications.Therefore, modification of the primary amine functionsto obtain crosslinking functionalities (e.g., methacry-lamides) is often performed.94–96,99–101 Processing oflow concentrations which are considered beneficialfor cell viability via direct SFF becomes difficult.4

However, in contrast to direct SFF, iRP has enabledthe construction of self-supporting scaffolds usingpolymer concentrations of 10 wt% and below(Fig. 8).3,95

Another applied biomaterial is silk fibroin, a silk-derived, protein-based material which is of specificinterest for the engineering of ligaments, bones andcartilage. It consists of amino-acids and is waterinsoluble, in contrast to gelatin. Conventional fabri-cation methods only allow limited control over theinternal architecture. Therefore, Liu et al. combinediRP with freeze drying to develop silk fibroin tissueengineering scaffolds with macro-channels and micro-pores thereby enabling optimal cell infiltration.50

Synthetic Polymers in Indirect Rapid Prototyping

Due to their mechanical properties and biodegrad-ability, polyesters are often applied for bone tissueengineering. These materials are often processed usingfused deposition modelling. However, this methoddoes not allow the creation of a secondary porestructure, for example by including porogens duringthe casting step. By using an indirect approach,

nanofibrous PLLA scaffolds, mimicking the morpho-logical functions of type I collagen, can be generatedby applying phase separation in dioxane.11,104 Phaseseparation can be combined with paraffin micro-spheres to generate a tri-porous structure.104 PCL is abiocompatible, semi-crystalline, slowly degradingpolymer. It combines flexibility with robust mechanicalproperties, which makes it an attractive candidate forbone tissue engineering.36,57,73 In addition, PCL hasalso been applied in eiRP for the generation of a car-otid artery. To this end, a dip coating process wascombined with salt leaching to create a porous con-struct.93 PLGA is another popular scaffold materialfor bone tissue engineering due to its robust mechan-ical properties..43,64,80

To demonstrate the obtainable resolution of iRP,Lee et al. fabricated villi-shaped scaffolds and the ef-fect of scaffold design on smooth muscle cell responsewas evaluated (Fig. 9). A higher cell density wasobserved for scaffolds with smaller villi-features (0.5 vs1 mm). This supports the hypothesis that a largerboundary surface area will better support mass trans-port.43 Another polymer which has been applied forbone tissue engineering is poly(propylene fumarate)(PPF), an unsaturated linear polyester, which incor-porates crosslinkable double bonds in its backbone.Furthermore, it has shown to support chondrocyte

FIGURE 8. Live/dead fluorescence images obtained of thescaffolds seeded with human foreskin fibroblasts (HFF) forboth types of scaffolds after 1 and 5 days of cell cultureincluding control image on tissue culture plastic. (Scale barsrepresent 200 lm) Reprinted from source Ref. 95 With per-mission of Springer.

Indirect Rapid Prototyping 71

growth combined with matrix deposition. Rapid pro-totyping of unconventional polymers often requirestedious optimization of experimental processingparameters, which can be surpassed by using iRP.

Poly-3-hydroxyoctanoate-co-3-hydroxyhexanoateand poly-4-hydroxybutyrate have been combined tocreate functional heart valves using the eiRP approach,which allows the use of porogens85,86 These materialsare thermoplastic elastomers which are produced via afermentation process.

Poly(2-oxazolines),a versatile material class withgreat potential for biological applications, have beenapplied in combination with MEW to generate aconstruct with microchannels.26

Ceramics in Indirect Rapid Prototyping

Porous ceramics are generally applied as scaffoldingmaterials for bone defects and spinal fusion applica-tions. In this respect, iRP is very appealing since anyorganic mould can be removed during sintering orpyrolysis, thereby elegantly combining two processingsteps. In addition, direct rapid prototyping of ceramicslurries is challenging and often requires a long opti-mization process. For bone tissue engineering, calciumphosphate is most often used under the form of tri-

calcium phosphate (TCP), a calcium salt of phosphoricacid. This material is commercially available for its usein medical and dental applications. To fill the iRPmould, TCP particles are combined with a bindersolution to obtain a ceramic slurry. After casting, thebinder (and mould) are removed via sintering.15,47

To increase the scaffold potential, growth factors suchas bone-morphogenic proteins (BMP) can be includedin the scaffold post-sintering. Especially of interest isBMP-2, a bone matrix protein stimulating mesenchy-mal cell differentiation into chondrocytes andosteoblasts.15

A second important ceramic material for tissueengineering purposes is hydroxyapatite (HA), a calciumphosphate and major component of human bones andteeth. It has been studied extensively as bone graftmaterial, since it enables the culture of many types ofosteogenic and bone marrow cells. Implantation studieshave already shown its ability to support ingrowth ofbone into porous implants35,36,38,46,52,69,96,99–101,104

(Fig. 10). Often, a thermally crosslinked binder is usedas a carrier for the HA particles, which can be removedduring sintering.13,14,30,32,75,106 However, if instead of abinder a self-setting solution is used, sintering is notrequired and growth factors such as BMP-2 can beincluded in the scaffold during fabrication.10

FIGURE 9. a–c, Macroscopic images of scaffolds with villi architecture with different dimensions. d–f, Cryosectional images. (leftto right: 0.5, 1, 0.5 mm; 0.5, 1, 1 mm; 1, 1, 1 mm; villus diameter; height; intervillus spacing respectively). Reprinted from sourceRef. 43 with permission from Wiley Periodicals, Inc.

HOUBEN et al.72

When comparing HA and TCP or blends thereof asscaffold material, HA or HA/TCP blends prove to besuperior to TCP in terms of cell viability, proliferationand osteogenic differentiation.81,82 In addition to theseconventional ceramic materials, alumina ceramicpowders have also been investigated as scaffold mate-rials.6,7

Composites and Blends in Indirect Rapid Prototyping

The casting process offers the possibility of creatingbiphasic and composite scaffolds which exhibit zoneswith different properties for structural tissue interfaceengineering. An interesting example is the combinationof ceramic and polymer composites. Material-relatedproperties such as flexibility of polyesters and tough-ness of ceramics can be combined as a composite or forthe creation of discrete zones within the scaffold.87

Inspired by natural bone, which is a composite of or-ganic and inorganic compounds, Li et al. combinedPLLA and chitosan with HA microspheres. PLLA isused for its high stability, while chitosan has a highbiocompatibility. The addition of HA microspheresincreases proliferation of pre-osteoblastic cells, result-ing in an ideal scaffold for bone tissue engineering(Fig. 11).44

By combining the synthetic polymer PPF, as de-scribed in paragraph 2.5.2 with b-TCP, an interestingcomposite can also obtained. In addition to its osteo-

conductivity, the mechanical properties of the materialcan increase during the early stages of degradation as aresult of continued crosslinking and by complexationof the polymeric carboxylic acid groups with calciumions.48

Besides biphasic scaffolds and blends, surfacetreatments can be applied to create multi-materialscaffolds. For example, chitosan and PCL mandibularcondyle shaped scaffolds were coated with apatite toincrease osteoinductive properties.41

APPLICATIONS OF INDIRECT RAPID

PROTOTYPING

Due to the versatile nature of the iRP method, andthe broad range of materials that can be applied,indirect scaffolds have been developed for differentapplications ranging from hard to soft tissues, tissueinterfaces and the creation of vascular networks(Fig. 12). By combining medical imaging with the 3Dscaffold production methods, patient-specific and de-fect-specific tissue engineering implants can be devel-oped. Furthermore, iRP has been applied for thecreation of microfluidic and vascular networks. How-ever, since this topic is somewhat different from con-ventional tissue engineering it is discussed in the‘‘Indirect Templates for the Formation of VascularNetworks’’ section.

FIGURE 10. A-D, Confocal laser scanning microscope image of a cellular network grown on HA scaffold. E, Overlay of image Bwith a virtual drawing of the scaffold sector. Reprinted from source Ref. 75 with permission from Wiley Periodicals, Inc

Indirect Rapid Prototyping 73

FIGURE 11. Biphasic PLA/HA ceramic scaffold (top, yellow: PLLA; bottom, blue: HA).Reprinted from source Ref. 44 with per-mission from Elsevier.

FIGURE 12. Applications of iRP in the engineering of hard tissue, soft tissue and tissue interfaces.

HOUBEN et al.74

Hard Tissue Engineering Using Indirect RapidPrototyping

The main advantage of iiRP is its rigorous controlin terms of scaffold architecture. As a result, the nat-ural morphology of bone tissue including the complexinternal microstructure with Volkman’s and Haversiancanals, can be carefully designed in the mould.10

Pore size and shape greatly influence the mechanicalproperties of a scaffold and tissue ingrowth. Increasedporosity will lead to a better ingrowth, but will de-crease the implant stiffness. Via design optimization,scaffold microstructures can be designed to match themechanical properties of the target tissue, whilemaintaining sufficient porosity for tissue ingrowth.14,30

In this respect, scaffolds with moduli and strengthvalues in the range of human trabecular bone (com-pressive modulus of 10–900 MPa, yield stress of 0.2–14 MPa) could be generated using a casting method.80

In addition to the mechanical properties, pore shapecan also influence cell growth. By choosing differentpore shapes this influence can be assessed. Channelsurface area and local curvature have been shown tostrongly influence the tissue growth rate. Roundchannels result in uniform growth of MC3T3-E1 cells.However in polygonal pores, tissue growth typicallystarts in the corners, whereas cells on the faces initiallyremained in a resting state. As a result, round pores areobtained over time regardless of the initial poreshape74 (Fig. 13). However, for identically shapedpores with a size range of 350–800 lm, no significant

effect on tissue ingrowth was observed in PCL evenafter 8 weeks of implantation.57

Detsch et al. compared bone marrow stromal cellcultivation on scaffolds generated via a direct dispenseplotting method and an iiRP method. They observedthat for the iiRP scaffolds a higher differentiation ofthe bone marrow stromal cells into precursor osteo-blasts occurred while the direct printed scaffoldexhibited a higher proliferation rate. They concludedthat both scaffold types were suitable for tissue cul-ture.20,21

One of the targeted tissues to be regenerated usingan iRP approach is craniofacial tissue. Based oncomputed tomography (CT) or magnetic resonanceimaging (MRI) scans the scaffold can closely mimic thearchitecture of the defect, for example the mandibularchondyle 31 or the zygoma.42 Human digit bone wasreconstructed by combining a phase separation tech-nique with an eiRP method including paraffinspheres.104

Aiming at cartilage tissue engineering an indirectPPF scaffold has been combined with a hyaluronicacid/collagen I hydrogel for cell delivery.45 Thesescaffolds combined load bearing qualities with en-hanced tissue regeneration. Both cubic and ellipsoidshaped pores resulted in good tissue infiltration with-out any significant difference.

Moore et al. developed a bio artificial graft for therepair of spinal cord injuries. Tailoring to the complexnature of these injuries they developed an implant ac-

FIGURE 13. Tissue formation depending on pore structure. Because of different cellular growth rates of cells located in thecorner or on the face of the pore, a final round shape is formed regardless of the initial pore morphology. Reprinted withpermission from The Royal Society publishing.74

Indirect Rapid Prototyping 75

tive both at the molecular, the cellular and the tissuelevel. An implant closely mimicking the morphology ofthe human spinal cord was fabricated using PLLA.Due to its specific architecture the scaffold could beseeded with multiple cell types mirroring the anatom-ical tissue locations. Schwann-cell loaded scaffoldswere implanted in vivo and promoted axon regenera-tion (Fig. 14).64

Soft Tissue Engineering Using Indirect RapidPrototyping

Using iRP, complex structures with high resolutionscan be achieved, which is demonstrated by printedscaffolds exhibiting a small villi architecture.40,42,43

Another complex tissue that has been generatedusing iRP is the heart valve. The design needs to beboth functional as well as physiologically correct and

was obtained based on x-ray computed tomography.The designs included human pulmonary and aortichomografts and were based on moulds fabricatedusing stereolithography followed by thermoforming ofporous PPF sheets to obtain the final structure.85,86

By combining eiRP with dip coating and saltleaching, porous, patient-specific, bifurbicating bloodvessels have been fabricated based on CT scans of acarotid artery (Fig. 15) which exhibited a mechanicalstrength within the range of natural human bloodvessels (1–3 MPa). The biocompatibility of the scaffoldwas confirmed using human umbilical vein endothelialcells (HUVECs).93

Indirect Rapid Prototyping as a Solution for TissueInterfaces

The treatment of periodontal diseases involves theformation of new cementum, periodontal ligament andalveolar bone. This challenging multi-tissue interfaceinspired the combination of a biphasic scaffold withPGA as periodontal ligament and PCL to engineer thebone compartment. A mould was developed to per-fectly fit a human tooth dentin slice attached to thePGA region of the scaffold (Fig. 16).72,73

Indirect Templates for the Formation of VascularNetworks

Vascularized Tissue Engineering Constructs

Vascularization in tissue engineering constructs re-mains a key challenge as many different cell types needa constant supply of oxygen and nutrients. In the ab-sence of blood vessels and capillaries, the thickness ofmost engineered tissues is limited to 100–200 lm.33

This need is partially resolved by creating porousstructures using (direct) rapid prototyping techniquesor other pore-creating methods (‘‘Generation ofMicropores Inside the Scaffold’’ section). However,even with these porous scaffolds nutrient and oxygensupply is driven by passive diffusion. As a result,achieving cell densities close to physiological popula-tions is extremely challenging since often necrotic areasare formed.62 The use of sacrificial templates offers anelegant alternative to conventional porous structuresvia the generation of multiscale, vascular-like channelsin engineered tissue, which enables active transport ofnutrients and waste.

Natural vascularised tissues consist of three maincomponents being the vascular lumen, the vascularwall lined with endothelial cells and a matrix of cellswhich surround the vascular channels. After removalof the vascular template the obtained channels can beperfused with endothelial cells to deposit a monolayer

FIGURE 14. Histological images of tissue formed within thespinal cord scaffold, segmented axon bundles are clearlyvisible. Reprinted with permission from Elsevier64.

HOUBEN et al.76

of cells. Consequently, hybrid cellularized scaffolds canbe obtained exhibiting the presence of different celltypes and a vascular network which closely mimicsnatural tissue.3,23,62

To enable template removal in the presence ofencapsulated cells template materials have to be care-fully selected taking into account a number ofrequirements. They cannot be cytotoxic and need to beremoved under mild conditions without relying oncytotoxic organic solvents. Examples of these materialsinclude agarose,3 gelatin23 and carbohydrate-glass.62

These can all be dissolved by melting or dissolution inwater or culture medium. An interesting approach toprevent leachables from entering into the matrix dur-ing template removal is the coating of vascular chan-nels prior to casting. During selective dissolution thisthin film prevents the leachables to enter the cellencapsulated hydrogel matrix while maintaining dif-fusion properties for biomolecules to the cells(Fig. 17).

Microfluidics

Complex microfluidic channels are of interest for arange of applications in biotechnology, in sensors, labon chips, organs on chips, chemical reactors and fluidicbased computers.91 They can be directly prototyped by(soft) lithography,28,52,76 laser ablation,46 deep protonwriting,102 micro-milling25 and ultra-precision dia-mond tooling.17 Also, they can be produced at low costin high volumes through replication techniques such asmicro-injection moulding,22 hot embossing18 or roll-to-roll printing.103 However, with these techniques thecreation of bifurcating true three-dimensional net-works remains challenging. Direct writing with fugitiveink offers interesting perspectives in the creation ofthese microfluidic channels. For example, the creationof biomimetic vascular networks to study fluid trans-port efficiency has already been reported.108

Therriault et al. have used direct-write assemblywith a fugitive ink to create channels with diameters

FIGURE 15. Bifurbicating carotid artery, reconstruction (a) and macroscopic image of the PCL scaffold (b). SEM images of thesurface (c and d).Reprinted from source Ref. 93 with permission from Elsevier.

FIGURE 16. Polymer casted, hybrid scaffold for engineering the human periodontal tooth-ligament (PDL) interface. Left: 3Ddesign, right: l-CT images and 3D reconstruction. Scale bar is 50 lm. Reused from source Ref. 72 with permission from Elsevier.

Indirect Rapid Prototyping 77

ranging from 10 to 300 lm. An epoxy resin was castand the ink was liquefied and removed forming a well-designed 3D epoxy network. By filling the channelswith a photopolymerizable resin and selectivelyblocking some channels by using a filter before UV-curing, a tower network was created to perform 1D,2D and 3D mixing experiments (Fig. 18).

CONCLUSION AND OUTLOOK

The application of iRP has proven to be a tooltackling limitations encountered during the applicationof direct SFF techniques. Well-defined 3D scaffoldscan be produced from biomaterials with mismatchedprocessing properties, from thermally unstable materi-

FIGURE 17. (a), Vascular unit cell. (b), Patterned vascular channels supporting human blood flow. Scale bars: 1 mm (left) 2 mm(right). (c), Encapsulated 10T1/2 cells in a filbrin gel, seeded HUVECs expressing mCherry. (d), Two intersecting channelsdemonstrating endothelisation. (e), Cross-section of a representative channel after 9 days in culture. (f), Single and multicellularsprouts formed by endothelial cells (arrowheads) Reprinted with permission from Macmillan Publishers Ltd: Nature Materials,62

copyright (2012).

HOUBEN et al.78

als, ceramics and composites. It allows a straightfor-ward combination of different materials within thesame scaffold, via blending of materials, or via thecreation of multi-zone materials. This has interestingapplications in the engineering of tissue interfaces.Material usage is highly efficient as only the volume ofthe final construct is required for scaffold productionwhereas most direct SFF techniques require excessiveamounts of material. Furthermore, the combination ofindirect rapid prototyping with conventional poreforming methods allows for the generation of struc-tures with a hierarchical porosity. As a result, superiorporosities can be obtained in comparison to direct SFFtechniques. Therefore, indirect iRP paves the way tomore complex constructs and opens up unprecedentedopportunities.

In applications where vascularization is a key issue,iRP offers unprecedented possibilities compared to thecurrently applied approaches. In this respect, thecombination of cell encapsulation with a vascularnetwork obtained via iRP perfused with endothelial

cells can mimic natural tissue more closely than everbefore. The application of cells encapsulated inhydrogels is of paramount importance. Therefore, weanticipate that more focus will be put on this elegantapproach to directly combine cells and materials in onefinal construct. Looking at the current applications foriRP it is striking that the majority of all scientific workin this field relates to hard tissue engineering. This hasa historical cause as hard tissue engineering applica-tions have always received a dominant attention in thebiomedical field. In addition, it can be attributed to theprocessing limitations experienced for ceramic materi-als often used in hard tissue engineering. Using iRP,these processing challenges are easily surpassed whichmakes this methodology highly attractive. Comparedto hard tissue engineering, soft tissue engineering re-ceives less attention but is considered an importantemerging application domain of iRP. We anticipatethat in the following decade novel applications andmaterials will find their way to the field of iRP.Therefore, the potential application ranges will further

FIGURE 18. a–c, Schematic representations and fluorescent microscope images of 1D, 2D and 3D microfluidic mixing experi-ments. Arrows indicate flow direction. Reprinted with permission from Macmillan Publishers Ltd: Nature Materials,91 copyright(2003).

Indirect Rapid Prototyping 79

be expanded also covering a higher number of softtissue engineering and tissue interface applications. Inthis respect, iRP has already proven to be an idealtechnique for the generation of scaffolds combiningmultiple materials thereby mimicking the complexinterface between hard and soft tissue. Furthermore,even outside the field of TE iRP can be applied forexample in the study of microfluidic processes. Inconclusion, the technique offers additional advantagesand new opportunities in areas where the conventionalrapid prototyping methods fail.

ACKNOWLEDGMENTS

The authors would like to acknowledge the financialsupport of Ghent University, the UGent Multidisci-plinary Research Partnership Nano-and Biophotonics,the Vrije Universiteit Brussel and Methusalem. JasperVan Hoorick and Sandra Van Vlierberghe would liketo acknowledge the Research Foundation Flanders(FWO, Belgium) for financial support under the formof respectively a PhD Grant and several ResearchGrants.

REFERENCES

1Arcaute, K., B. K. Mann, and R. B. Wicker. Stere-olithography of three-dimensional bioactive poly(ethyleneglycol) constructs with encapsulated cells. Ann. Biomed.Eng. 34:1429–1441, 2006.2Barry, R. A., R. F. Shepherd, J. N. Hanson, R. G. Nuzzo,P. Wiltzius, and J. A. Lewis. Direct-write assembly of 3Dhydrogel scaffolds for guided cell growth. Adv. Mater.21:2407–2410, 2009.3Bertassoni, L. E., M. Cecconi, V. Manoharan, M. Nik-khah, J. Hjortnaes, A. L. Cristino, G. Barabaschi, D.Demarchi, M. R. Dokmeci, Y. Yang, and A.Khademhosseini. Hydrogel bioprinted microchannel net-works for vascularization of tissue engineering constructs.Lab Chip 14:2202–2211, 2014.4Billiet, T., E. Gevaert, T. De Schryver, M. Cornelissen,and P. Dubruel. The 3D printing of gelatin methacry-lamide cell-laden tissue-engineered constructs with highcell viability. Biomaterials 35:49–62, 2014.5Billiet, T., M. Vandenhaute, J. Schelfhout, S. Van Vlier-berghe, and P. Dubruel. A review of trends and limita-tions in hydrogel-rapid prototyping for tissue engineering.Biomaterials 33:6020–6041, 2012.6Bose, S., J. Darsell, M. Kintner, H. Hosick, and A.Bandyopadhyay. Pore size and pore volume effects onalumina and TCP ceramic scaffolds. Mater. Sci. Eng. C23:479–486, 2003.7Bose, S., S. Suguira, and A. Bandyopadhyay. Processingof controlled porosity ceramic structures via fused depo-sition. Scr. Mater. 41:1009–1014, 1999.

8Brown, T. D., P. D. Dalton, and D. W. Hutmacher. Di-rect Writing by way of melt electrospinning. Adv. Mater.23:5651–5657, 2011.9Chang, P.-C., B.-Y. Liu, C.-M. Liu, H.-H. Chou, M.-H.Ho, H.-C. Liu, D.-M. Wang, and L.-T. Hou. Bone tissueengineering with novel rhBMP2-PLLA composite scaf-folds. J. Biomed. Mater. Res. A 81:771–780, 2007.

10Chen, Z. Z., D. C. Li, B. H. Lu, Y. P. Tang, M. L. Sun,and Z. Wang. Fabrication of artificial bioactive boneusing rapid prototyping. Rapid Prototyp. J. 10:327–333,2004.

11Chen, V. J., L. A. Smith, and P. X. Ma. Bone regenerationon computer-designed nano-fibrous scaffolds. 27:3973–3979, 2006; (Image (Table 2) was reprinted with permis-sion from Springer).

12Chia, H. N., and B. M. Wu. Recent advances in 3Dprinting of biomaterials. J. Biol. Eng. 9:4, 2015.

13Chu, T. M. G., J. W. Halloran, S. J. Hollister, and S. E.Feinberg. Hydroxyapatite implants with designed internalarchitecture. J. Mater. Sci. Mater. Med. 12:471–478, 2001.

14Chu, T. M. G., D. G. Orton, S. J. Hollister, S. E. Fein-berg, and J. W. Halloran. Mechanical and in vivo per-formance of hydroxyapatite implants with controlledarchitectures. Biomaterials 23:1283–1293, 2002; (Image(Table 1) was reprinted with permission from Elsevier).

15Chu, T.-M. G., S. J. Warden, C. H. Turner, and R. L.Stewart. Segmental bone regeneration using a load-bear-ing biodegradable carrier of bone morphogenetic protein-2. Biomaterials 28:459–467, 2007.

16De Beer, N., and A. Van Der Merwe. Patient-specificintervertebral disc implants using rapid manufacturingtechnology. Rapid Prototyp. J. 19:126–139, 2013.

17De Coster, D., D. Loterie, H. Ottevaere, M. Vervaeke, J.Van Erps, J. Missine, and H. Thienpont. Free-form opticsenhanced confocal Raman spectroscopy for optofluidiclab-on-chips. IEEE J. Sel. Top. Quantum Electron.21:2701108, 2015.

18De Coster, D., H. Ottevaere, M. Vervaeke, J. Van Erps,M. Callewaert, P. Wuytens, S. H. Simpson, S. Hanna, W.De Malsche, and H. Thienpont. Mass-manufacturablepolymer microfluidic device for dual fiber optical trap-ping. Opt. Express 23:30991, 2015.

19De Maria, C., A. De Acutis, and G. Vozzi. Indirect RapidPrototyping for Tissue Engineering. Elsevier, 2015. doi:10.1016/B978-0-12-800972-7/00008-6.

20Deisinger, U., S. Hamisch, M. Schumacher, F. Uhl, R.Detsch, and G. Ziegler. Fabrication of tailored hydrox-yapatite scaffolds: comparison between a direct and anindirect rapid prototyping technique. Key Eng. Mater.II(915–918):361–363, 2008.

21Detsch, R., F. Uhl, U. Deisinger, and G. Ziegler. 3D-Cultivation of bone marrow stromal cells on hydroxyap-atite scaffolds fabricated by dispense-plotting and nega-tive mould technique. J. Mater. Sci. Mater. Med. 19:1491–1496, 2008.

22Gadegaard, N., S. Mosler, and N. B. Larsen. Biomimeticpolymer nanostructures by injection molding. Macromol.Mater. Eng. 288:76–83, 2003.

23Golden, A. P., and J. Tien. Fabrication of microfluidichydrogels using molded gelatin as a sacrificial element.Lab Chip 7:720, 2007; (Image (Table 1) was reprinted withpermission from Elsevier).

24Graulus, G.-J., A. Mignon, S. Van Vlierberghe, H.Declercq, K. Feher, M. Cornelissen, J. C. Martins, and P.Dubruel. Cross-linkable alginate-graft-gelatin copolymers

HOUBEN et al.80

for tissue engineering applications. Eur. Polym. J. 2015.doi:10.1016/j.eurpolymj.2015.06.033.

25Grumann, M., J. Steigert, L. Riegger, I. Moser, B. En-derle, K. Riebeseel, G. Urban, R. Zengerle, and J. Ducree.Sensitivity enhancement for colorimetric glucose assays onwhole blood by on-chip beam-guidance. Biomed.Microdevices 8:209–214, 2006.

26Haigh, J. N., Y. Chuang, B. Farrugia, R. Hoogenboom,P. D. Dalton, and T. R. Dargaville. Hierarchically struc-tured porous poly(2-oxazoline) hydrogels. Macromol.Rapid Commun. 2015. doi:10.1002/marc.201500495.

27Hattiangadi, A., and A. Bandyopadhyay. Processing,Characterization and Modeling of Non-Random PorousCeramic Structures. Int. Solid Free. Fabr. Symp. 319–326,1999.

28Higgins, J. M., D. T. Eddington, S. N. Bhatia, and L.Mahadevan. Sickle cell vasoocclusion and rescue in amicrofluidic device. Proc. Natl. Acad. Sci. USA104:20496–20500, 2007.

29Ho, Y.-C., F.-M. Huang, and Y.-C. Chang. Cytotoxicityof formaldehyde on human osteoblastic cells is related tointracellular glutathione levels. J. Biomed. Mater. Res. BAppl. Biomater. 83:340–344, 2007.

30Hollister, S. J., R. A. Levy, T. M. Chu, J. W. Halloran,and S. E. Feinberg. An image-based approach fordesigning and manufacturing craniofacial scaffolds. Int. J.Oral Maxillofac. Surg. 29:67–71, 2000.

31Hollister, S., C. Lin, E. Saito, C. Lin, R. Schek, J. Taboas,J. Williams, B. Partee, C. Flanagan, A. Diggs, E. Wilke,G. Van Lenthe, R. M. Ller, T. Wirtz, S. Das, S. Feinberg,and P. Krebsbach. Engineering craniofacial scaffolds.Orthod. Craniofac. Res. 8:162–173, 2005.

32Hollister, S. J., R. D. Maddox, and J. M. Taboas. Optimaldesign and fabrication of scaffolds to mimic tissue prop-erties and satisfy biological constraints. Biomaterials23:4095–4103, 2002.

33Jain, R. K., P. Au, J. Tam, D. G. Duda, and D. Fuku-mura. Engineering vascularized tissue. Nat. Biotechnol.23:821–823, 2005.

34Kang, H.-W., and D.-W. Cho. Development of an indirectstereolithography technology for scaffold fabrication witha wide range of biomaterial selectivity. Tissue Eng. CMethods 18:719–729, 2012.

35Kang, K. H., L. A. Hockaday, and J. T. Butcher. Quan-titative optimization of solid freeform deposition ofaqueous hydrogels. Biofabrication 5:035001, 2013.

36Kang, H., J. Rhie, and D. Cho. Development of a bi-porescaffold using indirect solid freeform fabrication based onmicrostereolithography technology. Microelectron. Eng.86:941–944, 2009.

37Lam, C. X., X. Mo, S. Teoh, and D. Hutmacher. Scaffolddevelopment using 3D printing with a starch-based poly-mer. Mater. Sci. Eng. C 20:49–56, 2002.

38Landers, R., A. Pfister, U. Hubner, H. John, R. Sch-melzeisen, and R. Milhaupt. Fabrication of soft tissueengineering scaffolds by means of rapid prototypingtechniques. J. Mater. Sci. 37:3107–3116, 2002.

39Langer, R., and J. P. Vacanti. Tissue engineering. Science(80-) 260:920–926, 1993.

40Lee, J.-Y., B. Choi, B. Wu, and M. Lee. Customizedbiomimetic scaffolds created by indirect three-dimensionalprinting for tissue engineering. Biofabrication 5:045003,2013.

41Lee, J.-Y., B. Choi, B. Wu, and M. Lee. Customizedbiomimetic scaffolds created by indirect three-dimensional

printing for tissue engineering. Biofabrication 5:12–14,2013.

42Lee, M., J. C. Y. Dunn, and B. M. Wu. Scaffold fabri-cation by indirect three-dimensional printing. Biomaterials26:4281–4289, 2005.

43Lee, M., B. M. Wu, and J. C. Y. Dunn. Effect of scaffoldarchitecture and pore size on smooth muscle cell growth.J. Biomed. Mater. Res. A 87A:1010–1016, 2008.

44Li, L. H., K. P. Kommareddy, C. Pilz, C. R. Zhou, P.Fratzl, and I. Manjubala. In vitro bioactivity of biore-sorbable porous polymeric scaffolds incorporatinghydroxyapatite microspheres. Acta Biomater. 6:2525–2531, 2010.

45Liao, E., M. Yaszemski, P. Krebsbach, and S. Hollister.Tissue-engineered cartilage constructs using compositehyaluronic acid/collagen I hydrogels and designedpoly(propylene fumarate) scaffolds. Tissue Eng. 13:537–550, 2007.

46Lim, D., Y. Kamotani, B. Cho, J. Mazumder, and S.Takayama. Fabrication of microfluidic mixers and artifi-cial vasculatures using a high-brightness diode-pumpedNd:YAG laser direct write method. Lab Chip 3:318–323,2003.

47Limpanuphap, S., and B. Derby. Manufacture of bio-materials by a novel printing process. J. Mater. Sci. Ma-ter. Med. 13:1163–1166, 2002.

48Lin, C.-Y., R. Schek, A. Mistry, X. Shi, A. G. Mikos, P.H. Krebsbach, and S. J. Hollister. Functional bone engi-neering using ex vivi gene therapy and topology-opti-mized, biodegradable polymer composite scaffolds. TissueEng. 11:1589–1598, 2005.

49Liska, R., F. Schwager, C. Maier, R. Cano-Vives, and J.Stampfl. Water-soluble photopolymers for rapid proto-typing of cellular materials. J. Appl. Polym. Sci. 97:2286–2298, 2005.

50Liu, M. J. J., S. M. Chou, C. K. Chua, B. C. M. Tay, andB. K. Ng. The development of silk fibroin scaffolds usingan indirect rapid prototyping approach: morphologicalanalysis and cell growth monitoring by spectral-domainoptical coherence tomography. Med. Eng. Phys. 35:253–262, 2013; (Image (Table 2) was reprinted with permissionfrom Springer.).

51Liu, C. Z., E. Sachlos, D. A. Wahl, Z. W. Han, and J. T.Czernuszka. On the manufacturability of scaffold mouldusing a 3D printing technology. Rapid Prototyp. J.13:163–174, 2007.

52Liu, R. H., M. A. Stremler, K. V. Sharp, M. G. Olsen, J.G. Santiago, R. J. Adrian, H. Aref, and D. J. Beebe.Passive mixing in a three-dimensional serpentinemicrochannel. J. Microelectromech. Syst. 9:190–197, 2000.

53Liu, C. Z., Z. D. Xia, Z. W. Han, P. A. Hulley, J. T.Triffitt, and J. T. Czernuszka. Novel 3D collagen scaffoldsfabricated by indirect printing technique for tissue engi-neering. J. Biomed. Mater. Res. B Appl. Biomater. 85:519–528, 2008.

54Lu, L., Q. Zhang, D. Wootton, R. Chiou, D. Li, B. Lu, P.Lelkes, and J. Zhou. Biocompatibility and biodegradationstudies of PCL/b-TCP bone tissue scaffold fabricated bystructural porogen method. J. Mater. Sci. Mater. Med.23:2217–2226, 2012; (Image (Table 1) was reprinted withpermission from Elsevier).

55Makitie, A. A., J. Korpela, L. Elomaa, M. Reivonen, A.Kokkari, M. Malin, H. Korhonen, X. Wang, J. Salo, E.Sihvo, M. Salmi, J. Partanen, K.-S. Paloheimo, J. Tuomi,T. Narhi, and J. Seppala. Novel additive manufactured

Indirect Rapid Prototyping 81

scaffolds for tissue engineered trachea research. ActaOtolaryngol. 133:412–417, 2013.

56Manjubala, I., A. Woesz, C. Pilz, M. Rumpler, N. Fratzl-Zelman, P. Roschger, J. Stampfl, and P. Fratzl. Biomi-metic mineral-organic composite scaffolds with controlledinternal architecture. J. Mater. Sci. Mater. Med. 16:1111–1119, 2005.

57Mantila Roosa, S. M., J. M. Kemppainen, E. N. Moffitt,P. H. Krebsbach, and S. J. Hollister. The pore size ofpolycaprolactone scaffolds has limited influence on boneregeneration in an in vivo model. J. Biomed. Mater. Res. A92:359–368, 2010.

58Markovic, M., J. Van Hoorick, K. Holzl, M. Tromayer,P. Gruber, S. Nurnberger, P. Dubruel, S. Van Vlier-berghe, R. Liska, and A. Ovsianikov. Hybrid tissueengineering scaffolds by combination of three-dimen-sional printing and cell photoencapsulation. J. Nan-otechnol. Eng. Med. 6:021004, 2015.

59Melchels, F. P. W., J. Feijen, and D. W. Grijpma. A re-view on stereolithography and its applications inbiomedical engineering. Biomaterials 31:6121–6130, 2010.

60Melican, M. C., M. C. Zimmerman, M. S. Dhillon, A. R.Ponnambalam, A. Curodeau, and J. R. Parsons. Three-dimensional printing and porous metallic surfaces: a neworthopedic application. J. Biomed. Mater. Res. 55:194–202, 2001.

61Meyer, U. The history of tissue engineering and regener-ative medicine in perspective. In: Fundamentals of TissueEngineering and Regenerative Medicine, edited by U.Meyer, Th Meyer, J. Handschel, and J. P. Wiesmann.Berlin: Springer, 2009, pp. 5–12.

62Miller, J. S., K. R. Stevens, M. T. Yang, B. M. Baker, D.-H. T. Nguyen, D. M. Cohen, E. Toro, A. A. Chen, P. A.Galie, X. Yu, R. Chaturvedi, S. N. Bhatia, and C. S.Chen. Rapid casting of patterned vascular networks forperfusable engineered three-dimensional tissues. Nat.Mater. 11:768–774, 2012.

63Mondrinos, M. J., R. Dembzynski, L. Lu, V. K. C.Byrapogu, D. M. Wootton, P. I. Lelkes, and J. Zhou.Porogen-based solid freeform fabrication of polycapro-lactone-calcium phosphate scaffolds for tissue engineer-ing. Biomaterials 27:4399–4408, 2006.

64Moore, M. J., J. A. Friedman, E. B. Lewellyn, S. M.Mantila, A. J. Krych, S. Ameenuddin, A. M. Knight, L.Lu, B. L. Currier, R. J. Spinner, R. W. Marsh, A. J.Windebank, and M. J. Yaszemski. Multiple-channelscaffolds to promote spinal cord axon regeneration. Bio-materials 27:419–429, 2006.

65Moroni, L., R. Schotel, D. Hamann, J. R. de Wijn, and C.A. van Blitterswijk. 3D fiber-deposited electrospun inte-grated scaffolds enhance cartilage tissue formation. Adv.Funct. Mater. 18:53–60, 2008.

66Nichol, J. W., S. T. Koshy, H. Bae, C. M. Hwang, S.Yamanlar, and A. Khademhosseini. Cell-laden micro-engineered gelatin methacrylate hydrogels. Biomaterials31:5536–5544, 2010.

67Ovsianikov, A., A. Deiwick, S. Van Vlierberghe, P. Du-bruel, M. Lena, G. Drager, and B. Chichkov. Laser fab-rication of three-dimensional CAD scaffolds fromphotosensitive gelatin for applications in tissue engineer-ing. Biomacromolecules 12:851–858, 2011.

68Ovsianikov, A., A. Deiwick, S. Van Vlierberghe, M.Pflaum, M. Wilhelmi, P. Dubruel, and B. Chichkov. Laserfabrication of 3D gelatin scaffolds for the generation ofbioartificial tissues. Materials (Basel) 4:288–299, 2011.

69Ovsianikov, A., S. Muhleder, J. Torgersen, Z. Li, X.-H.Qin, S. Van Vlierberghe, P. Dubruel, W. Holnthoner, H.Redl, R. Liska, and J. Stampfl. Laser photofabrication ofcell-containing hydrogel constructs. Langmuir 30:3787–3794, 2014.

70Ovsianikov, A., A. Ostendorf, and B. N. Chichkov.Three-dimensional photofabrication with femtosecondlasers for applications in photonics and biomedicine. Appl.Surf. Sci. 253:6599–6602, 2007.

71Park, S., G. Kim, Y. C. Jeon, Y. Koh, and W. Kim. 3Dpolycaprolactone scaffolds with controlled pore structureusing a rapid prototyping system. J. Mater. Sci. Mater.Med. 20:229–234, 2009.

72Park, C. H., H. F. Rios, Q. Jin, M. E. Bland, C. L.Flanagan, S. J. Hollister, and W. V. Giannobile. Biomi-metic hybrid scaffolds for engineering human tooth-liga-ment interfaces. Biomaterials 31:5945–5952, 2010.

73Park, C. H., H. F. Rios, Q. Jin, J. V. Sugai, M. Padial-molina, A. D. Taut, C. L. Flanagan, S. J. Hollister, andW. V. Giannobile. Tissue engineering bone-ligamentcomplexes using fiber-guiding scaffolds. Biomaterials33:137–145, 2012.

74Rumpler, M., A. Woesz, J. W. Dunlop, J. T. van Dongen,and P. Fratzl. The effect of geometry on three-dimen-sional tissue growth. J. R. Soc. Interface 5:1173–1180,2008.

75Rumpler, M., A. Woesz, F. Varga, I. Manjubala, K.Klaushofer, and P. Fratzl. Three-dimensional growthbehavior of osteoblasts on biomimetic hydroxylapatitescaffolds. J. Biomed. Mater. Res. 81A:40–50, 2007.

76Runyon, M. K., B. L. Johnson-Kerner, C. J. Kastrup, T.G. Van Ha, and R. F. Ismagilov. Propagation of bloodclotting in the complex biochemical network of hemostasisis described by a simple mechanism. J. Am. Chem. Soc.129:7014–7015, 2007.

77Sachlos, E., N. Reis, C. Ainsley, B. Derby, and J. T.Czernuszka. A process to make collagen scaffolds with anartificial circulatory system using rapid prototyping. Ma-ter. Res. Soc. Symp. Proc. p. 758, 2002.

78Sachlos, E., N. Reis, C. Ainsley, B. Derby, and J. T.Czernuszka. A Process to Make Collagen Scaffolds withan Artificial Circulatory System using Rapid Prototyping.Mater. Res. Soc. Symp. Proc., 2002.

79Sachlos, E., N. Reis, C. Ainsley, B. Derby, and J. T.Czernuszka. Novel collagen scaffolds with predefinedinternal morphology made by solid freeform fabrication.Biomaterials 24:1487–1497, 2003.

80Saito, E., H. Kang, J. M. Taboas, A. Diggs, C. L.Flanagan, and S. J. Hollister. Experimental and compu-tational characterization of designed and fabricated 50:50PLGA porous scaffolds for human trabecular boneapplications. J. Mater. Sci. Mater. Med. 21:2371–2383,2010; (Image (Table 2) was reprinted with permissionfrom Springer.).

81Schumacher, M., U. Deisinger, R. Detsch, and G. Ziegler.Indirect rapid prototyping of biphasic calcium phosphatescaffolds as bone substitutes: influence of phase compo-sition, macroporosity and pore geometry on mechanicalproperties. J. Mater. Sci. Mater. Med. 21:3119–3127,2010.

82Schumacher, M., F. Uhl, R. Detsch, U. Deisinger, and G.Ziegler. Static and dynamic cultivation of bone marrowstromal cells on biphasic calcium phosphate scaffoldsderived from an indirect rapid prototyping technique. J.Mater. Sci. Mater. Med. 21:3039–3048, 2010.

HOUBEN et al.82

83Seol,Y. J., J. Y.Kim,E.K. Park, S.Y.Kim, andD.W.Cho.Fabrication of a hydroxyapatite scaffold for bone tissueregeneration using microstereolithography and moldingtechnology.Microelectron. Eng. 86:1443–1446, 2009.

84Skoog, S. A., P. L. Goering, and R. J. Narayan. Stere-olithography in tissue engineering. J. Mater. Sci. Mater.Med. 25:845–856, 2013.

85Sodian, R., M. Loebe, A. Hein, D. P. Martin, S. P.Hoerstrup, E. V. Potapov, H. Hausmann, T. Lueth, andR. Hetzer. Application of stereolithography for scaffoldfabrication for tissue engineered heart valves. Am. Soc.Artif. Intern. Organs J. 48:12–16, 2002.

86Sodian, R., J. S. Sperling, D. P. Martin, A. Egozy, U.Stock, J. E. Mayer, and J. P. Vacanti. Fabrication of atrileaflet heart valve scaffold from a polyhydroxyalka-noate biopolyester for use in tissue engineering. TissueEng. 6:183–188, 2000.

87Taboas, J. M., R. D. Maddox, P. H. Krebsbach, and S. J.Hollister. Indirect solid free form fabrication of local andglobal porous, biomimetic and composite 3D polymer-ceramic scaffolds. Biomaterials 24:181–194, 2003.

88Tan, J. Y., C. K. Chua, and K. F. Leong. Indirect fabri-cation of gelatin scaffolds using rapid prototyping tech-nology. Virtual Phys. Prototyp. 5:45–53, 2010.

89Tan, J. Y., C. K. Chua, and K. F. Leong. Fabrication ofchanneled scaffolds with ordered array of micro-poresthrough microsphere leaching and indirect. Rapid Proto-typ. Tech. 15:83–96, 2013. doi:10.1007/s10544-012-9690-3;(Image (Table 1) was reprinted with permission fromElsevier).

90Therriault, D., R. F. Shepherd, S. R. White, and J. A.Lewis. Fugitive inks for direct-write assembly of three-dimensional microvascular networks. Adv. Mater. 17:395–399, 2005.

91Therriault, D., S. R. White, and J. A. Lewis. Chaoticmixing in three-dimensional microvascular networks fab-ricated by direct-write assembly. Nat. Mater. 2:265–271,2003.

92Torgersen, J., A. Ovsianikov, V. Mironov, N. Pucher, X.Qin, Z. Li, K. Cicha, T. Machacek, R. Liska, V. Jantsch,and J. Stampfl. Photo-sensitive hydrogels for three-di-mensional laser microfabrication in the presence of wholeorganisms. J. Biomed. Opt. 17:105008, 2012.

93Uchida, T., S. Ikeda, H. Oura, M. Tada, T. Nakano, T.Fukuda, T. Matsuda, M. Negoro, and F. Arai. Devel-opment of biodegradable scaffolds based on patient-specific arterial configuration. J. Biotechnol. 133:213–218,2008.

94Van Den Bulcke, A. I., B. Bogdanov, N. De Rooze, E. H.Schacht, M. Cornelissen, and H. Berghmans. Structuraland rheological properties of methacrylamide modifiedgelatin hydrogels. Biomacromolecules 1:31–38, 2000.

95Van Hoorick, J., H. Declercq, A. De Muynck, A. Hou-ben, L. Van Hoorebeke, R. Cornelissen, J. Van Erps, H.Thienpont, P. Dubruel, and S. Van Vlierberghe. Indirectadditive manufacturing as an elegant tool for the pro-duction of self-supporting low density gelatin scaffolds. J.Mater. Sci. Mater. Med. 26:247, 2015.

96Van Rie, J., H. Declercq, J. Van Hoorick, M. Dierick, L.Van Hoorebeke, R. Cornelissen, H. Thienpont, P. Du-bruel, and S. Van Vlierberghe. Cryogel-PCL combinationscaffolds for bone tissue repair. J. Mater. Sci. Mater.Med. 26:5465, 2015.

97Van Vlierberghe, S., V. Cnudde, P. Dubruel, B. Mass-chaele, A. Cosijns, I. De Paepe, P. J. S. Jacobs, L. VanHoorebeke, J. P. Remon, and E. Schacht. Porous gelatinhydrogels: 1. Cryogenic formation and structure analysis.Biomacromolecules 8:331–337, 2007.

98Van Vlierberghe, S., P. Dubruel, E. Lippens, B. Mass-chaele, L. Van Hoorebeke, M. Cornelissen, R. Unger, C.J. Kirkpatrick, and E. Schacht. Toward modulating thearchitecture of hydrogel scaffolds: curtains versus chan-nels. J. Mater. Sci. Mater. Med. 19:1459–1466, 2008.

99Van Vlierberghe, S., P. Dubruel, and E. Schacht.Biopolymer-based hydrogels as scaffolds for tissue engi-neering applications: a review. Biomacromolecules12:1387–1408, 2011.

100Van Vlierberghe, S., B. Fritzinger, J. C. Martins, and P.Dubruel. Hydrogel network formation revised: high-res-olution magic angle spinning nuclear magnetic resonanceas a powerful tool for measuring absolute hydrogel cross-link efficiencies. Appl. Spectrosc. 64:1176–1180, 2010.

101Van Vlierberghe, S., E. Schacht, and P. Dubruel. Rev-ersible gelatin-based hydrogels: finetuning of materialproperties. Eur. Polym. J. 47:1039–1047, 2011.

102Verschooten, T., H. Ottevaere, M. Vervaeke, J. Van Erps,and H. Thienpont. Proof-of-concept demonstration of atotal internal reflection based module for fluorescence andabsorbance detection using a 3D-printed syringe pump.Proc. SPIE 9130:91300E–91300E-11, 2014.

103Vig, A. L., T. Makela, P. Majander, V. Lambertini, J.Ahopelto, and A. Kristensen. Roll-to-roll fabricated lab-on-a-chip devices. J. Micromech. Microeng. 21:035006,2011.

104Wang, P., J. Hu, and P. X. Ma. The engineering of pa-tient-specific, anatomically shaped, digits. Biomaterials30:2735–2740, 2009.

105Wilson, C. E., J. D. De Bruijn, C. A. Van Blitterswijk, A.J. Verbout, W. J. A. Dhert, J. D. de Bruijn, C. A. vanBlitterswijk, A. J. Verbout, and W. J. A. Dhert. Designand fabrication of standardized hydroxyapatite scaffoldswith a defined macro-architecture by rapid prototyping forbone-tissue-engineering research. J. Biomed. Mater. Res. A68:1–10, 2003.

106Woesz, A., M. Rumpler, J. Stampfl, F. Varga, N. Fratzl-Zelman, P. Roschger, K. Klaushofer, and P. Fratzl. To-wards bone replacement materials from calcium phos-phates via rapid prototyping and ceramic gelcasting.Mater. Sci. Eng., C 25:181–186, 2005; (Image (Table 2)was reprinted with permission from Springer).

107Woodfield, T. B. F., J. Malda, J. de Wijn, F. Peters, J.Riesle, and C. A. van Blitterswijk. Design of porousscaffolds for cartilage tissue engineering using a three-di-mensional fiber-deposition technique. Biomaterials25:4149–4161, 2004.

108Wu, W., C. J. Hansen, A. M. Aragon, P. H. Geubelle, S.R. White, and J. A. Lewis. Direct-write assembly of bio-mimetic microvascular networks for efficient fluid trans-port. Soft Matter 6:739, 2010.

109Yang, S., K.-F. Leong, D. Zhaohui, and C. Chua. Reviewthe design of scaffolds for use in tissue engineering. Part II.Rapid prototyping techniques. Tissue Eng. 8:1–11, 2002.

110Yeong, W., C. Chua, K. Leong, M. Chandrasekaran, andM. Lee. Comparison of drying methods in the fabricationof collagen scaffold via indirect. Rapid Prototyp. 82:260–266, 2006. doi:10.1002/jbmb.

Indirect Rapid Prototyping 83