Three-dimensional printing with biomaterials in craniofacial and dental tissue engineering

Polymer biomaterials for craniofacial bone TE.

Fabricating maxillofacial bone scaffold is a major application of AM technology in craniofacial usage. The selection of an ideal bone graft material relies on multiple factors such as material viability, graft size, porosity, hydrophilic, biodegradability, osteoconductivity and osteoinductivity. It was first reported that synthetic polymeric materials could generate AM bone scaffolds. Many polymers are printable, for they often have proper melting ranges to fulfill the technique requirement of shaping with FDM or binder jetting. As far back as in 1996, PLA was used as AM material in computer aided design (CAD) bone generation (Giordano et al., 1996). After that, other polymeric scaffolds have been increasingly developed in AM techniques, such as PCL (Williams et al., 2005; Lohfeld et al., 2012; Korpela et al., 2013; Van Bael et al., 2013; Temple et al., 2014a), poly(lactic-co-glycolic acid) (PLGA) (Luangphakdy et al., 2013), poly(trimethylene carbonate) (PTMC) (Blanquer, Sharifi & Grijpma, 2012) and so on. As a widely used biomedical material, PLA has good biocompatibility as implants with FDA clearance. Printed PLA bars have physical properties of maximum measured tensile strength. The maximum measured tensile strength of low molecular weight PLLA (53 000) is 17.40 ± 0.71 MPa, while that of high molecular weight PLLA (312,000) is 15.94 ± 1.50 MPa (Giordano et al., 1996). PCL is an alternative with PLA because it does not release acid in PLA remodeling. This means it is more resistant in vivo. PCL also has a lower glass transition temperature and melting temperature, making it superior to PLA in certain bone grafting applications. For instance, PCL can be easily blended with other materials, including tricalcium phosphate (TCP), hydroxyapatite (HA) and bioactive glass (BAG), due to its low melting temperature (Korpela et al., 2013). In addition, the compressive module of PCL can be increased up to 30–40% by adding 10 wt % of BAG.

As modifications for the mechanical performances (Duan & Wang, 2010), polymers are also blended in defined ratios to make printable composites, such as PCL/PLGA by FDM (Shim et al., 2014) and PLGA/PVA by binder jetting (Ge et al., 2009). PVA also serves as a porogen in the printed architectures by taking advantage of its water-soluble properties. PVA-blended HA was printed by SLS to study the feasibility of composite scaffold (Simpson et al., 2008). SEM observations showed significant improvements in the sintering effects and to be a suitable material when processed by SLS for TE scaffolds.

Cells and animal models used in craniofacial bone TE.

The selection of cell is important for bone TE. For orthopedic and maxillofacial researches, primary stem cells as bone marrow stromal cells (BMSC) (Fedorovich et al., 2009; Rath et al., 2012) and adipose derived stem cells (ADSC) (Temple et al., 2014a) are wildly applied to seed cell types. Fibroblasts are used for viability test and proliferation essay, as well as human multi-potent dental neural crest-derived progenitor cells (dNC-PCs) (Fierz et al., 2008). Multiple bone cell lines are applied in AM studies, including MC3T3-E1 (Leukers et al., 2005; Khalyfa et al., 2007; Lan et al., 2009; Melchels, Feijen & Grijpma, 2010; Blanquer, Sharifi & Grijpma, 2012), SaOS-2 (Duan & Wang, 2010; Wang et al., 2013), C3H/10T1/2 cells (Inzana et al., 2014) and MG-63 (Feng et al., 2014a; Feng et al., 2014b). With osteogenic induction, the attached bone cells not only exhibited cell viability around 60%–90%, but also kept potential of osteogenic differentiation which is confirmed by observing bone metabolism related RNA and protein expression, such as runt-related transcription factor 2 (RUNX2), bone morphogenetic proteins (BMPs), alkaline phosphatase (ALP) and osteonectin (ON) activity. For cells used in craniofacial bone TE, there are different advantages for different cells. Bone cell lines as MC3T3-E1, SaOS-2, c3h/10T1/2, MG-63 were often used for initial screening of biological activity of materials (Przekora, 2019). Since these cells are tumor-derived cell lines or immortalized osteoblast cell lines, their gene expressions are quite different from those of primary cells (Pautke et al., 2004). The best seed cells for craniofacial bone TE are still considered to be primary OBs because of their behavior in studying osteoconductive and osteopromotive properties (Przekora, 2019). The advantage of using stem cells also include testing the osteoconductive ability of printing materials (Temple et al., 2014b). What’s more, many kinds of tissue can be the source of autologous stem cells.

Several animals had been taken in AM mandible scaffold research. Rabbits are most frequently used in the study of mandibular bone repair (Alfotawei et al., 2014). A protocol described the usage of three-dimensional printed scaffolds with multipotent mesenchymal stromal cell (MSCs) in mandibular reconstruction of rabbits. They used BMSC and ADSC from rabbits (Fang et al., 2017). One of the previous studies was performed on six mature minipigs (Fig. 2). The researchers created four mandibular defects on each pig. After the defect sites were modelled by CAD/CAM techniques, scaffolds with complex geometries and very fine structures were produced by AM technology. Then the autologous porcine bone cells were seeded on these polylactic acid/polyglycolic acid (PLA/PGA) copolymer scaffolds. Implanting these tissue-constructs into the bone defects supported bone reconstruction (Meyer, Neunzehn & Wiesmann, 2012). What’s more, in a recent study, researchers proved that the craniofacial reconstruction including mandible could be achieved through 3D bioprinting. They presented an integrated tissue-organ printer (ITOP) that can fabricate stable, human-scale tissue constructs of any shape. They also found vascularized bone growth in the central and peripheral portion in vivo trails of rats (Kang et al., 2016). For periodontal bone regeneration, at least 4 mm augmentations of craniofacial bone had already been achieved with synthetic monetite blocks. 3D printing TCP plates were used as onlay grafts in periodontal surgery. The 4.0- and 3.0-mm high blocks were filled with newly formed bone with 35% and 41% of respective volumes (Torres et al., 2011a; Torres et al., 2011b). These 3D-printed customized synthetic onlay grafts were further used in dental implant surgery to achieve bone augments (Tamimi et al., 2014). Direct writing (DW) technology had been applied to produce a TCP scaffolds to repair the rabbit trephine defect. The scaffolds had micropores ranging from 250 × 250 µm up to 400 × 400 µm. After 16 weeks, 30% of the scaffold was remodeled by osteoclast activity with new bone filling in the scaffolds and across the defects (Ricci et al., 2012). These studies suggested that AM scaffold with tissue engineering could be used in human craniofacial defect repair in the future.

Technique challenges for craniofacial bone printing and current strategies.

Although cell migration and proliferation inside the porous scaffold were observed in an AM HA scaffolds with inner-connective pores (Fierz et al., 2008), for all the porous scaffolds, it is still a big challenge to keep good cell viability in the central area. Insufficient nutrition and oxygen in static culture lead to cell necrosis and make low cell density area. The method of dynamic cultivation can partly solve this problem. A dynamic cultivation system by perfusion containers strongly increased the MC3T3-E1 population compared to the static cultivation method in a 7-day in vitro cultivation. Close contact between cells and HA granules were observed deeply in the printed structure (Leukers et al., 2005). In another study, application of perfusion bioreactor system to a BCP binder jetting fabricated scaffold not only successfully reversed the decreased OB and BMSC cell numbers but also increased their differentiation potential (Rath et al., 2012).

Incomplete healing is another current limitation to AM bone grafts. Therefore, growth factors are applied in scaffolds. Bone morphology protein-2 (BMP2), a bone growth factor with strong bone induction property, is often used. The controlled release of BMP2 can be achieved by surface coating or nanoparticles embedding. More consideration is required according to the printing procedure for AM scaffolds. BMP2 loaded gelatin microparticles (GMPs) was used as a sustained release system and dispersed in hydrogel-based constructs, comparing with direct inclusion of BMP2 in alginate or control GMPs (Poldervaart et al., 2013). In another study with a multi-head deposition system (MHDS) , rhBMP2 was loaded by either gelatin (for short-term delivery within a week) or collagen (for long-term delivery up to 28 days) and dispensed directly into the hollow microchannel structure of PCL/PLGA scaffold during the printing process (Shim et al., 2014). The in vivo micro-computed tomography (micro CT) and histological analyses indicated that CL/PLGA/collagen/rhBMP2 scaffolds lead to superior bone healing quality at both 4 and 8 weeks, without inflammatory response. Transforming growth factor-β (TGF-β) was another important growth factor widely used in osteoblast differentiation and animal models (Nikolidakis et al., 2009).

Due to the hydrophobic feature of most printable materials, surface modification can be exploited to improve biocompatibility. Collagen is a widely used coating material for AM bone scaffold coating. The flexural strength and toughness of a calcium phosphate scaffold was significantly improved by coating a 0.5 wt% collagen film (Inzana et al., 2014). Biomimetic and β-TCP (Luangphakdy et al., 2013) can enhance the surface roughness and increase bone differentiation, thus may minimizing the need for expensive bone growth factors (Gibbs et al., 2014) (Table 1).

Polymer biomaterials for craniofacial cartilage TE.

Cartilage is one of the few tissues that are not vascularized, which makes its regeneration unique. The most widely applied techniques in cartilage printing included FDM, SLA and SLS. For cartilage repair, polymeric materials like PLA, PCL as well as PLGA were most common cartilage scaffolds. Another kind of major material was the hydrogel. Hydrogel could mimic the elastic module of cartilage and have been applied for cartilage reparation for a long time. Recent study showed PEG hydrogel had promising potential for cartilage bioprinting (Cui et al., 2012).

Cells for craniofacial cartilage TE in AM approaches.

Chondrocytes were the standard seed cells in cartilages TE, but chondrocytes from different cartilage subtypes exhibited different differentiation. In AM cartilage regeneration, to generate different cartilage subtypes, chondrocytes were harvested from several kinds of cartilages. In one research, rib cartilage cells were co-cultured with adherent stromal cells in a porous PCL scaffolds fabricated by FDM, making a culture system which may have potential of clinical usage (Cao, Ho & Teoh, 2003). In one research, porcine articular chondrocytes were seeded in PLGA scaffold fabricated with liquid-frozen deposition manufacturing, cultured for a total of 28 days. Final results showed that cells proliferated well and secreted abundant extracellular matrix (Yen et al., 2009). Not only chondrocytes, but also stem cells were also applied in cartilages TE, such as MSCs and so on (Pati et al., 2015). Interestingly, bone marrow clots (MC) as a promising resource proved to be a highly efficient, reliable, and simple cell resource that improved the biological performance of scaffolds as well. The FDM printed PCL-HA scaffold incubated with MC exhibited significant improvements in cell proliferation and chondrogenic differentiation. This study suggested that 3D printing scaffolds, MC could provide a promising candidate for cartilage regeneration (Yao et al., 2015). Stem cell-based approach and chondrocyte-based approach were common choices for cartilage regenerations. The major advantage of using stem cells is that autologous transplantation can be implemented (Walter, Ossendorff & Schildberg, 2019). Unlike chondrocytes, autologous stem cells, such as BMSCs or ADSCs, are rich in source. Xenografts of chondrocytes is not a good choice for human cartilage repair for there are immunological reactions (Stone et al., 1997). It is also reported that chondrocytes lost the chondrogenic differentiation after several passages (Von der Mark et al., 1977; Frohlich et al., 2007). On the other hand, the stem cells may form fibrocartilage-like tissue in defect without grows factors (Yoshioka et al., 2013). Differences in depth of the defect also affect the cartilage regeneration, which should be selected according to research purposes (Nixon et al., 2011).

AM application for TMJ cartilage.

The temporal mandibular joint (TMJ) disc is a heterogeneous fibrocartilaginous tissue which plays a vital role in its function. It was reported recently that researchers had developed TMJ disc scaffold with spatiotemporal delivery of connective tissue growth factor (CTGF) and transforming growth factor beta 3 (TGFβ3) which induced fibrochondrogenic differentiation of MSCs. They used layer-by-layer deposition printing technique with polycaprolactone (PCL) to fabricate the scaffold. CTGF and TGFβ3 were used as growth factors and human MSCs were used as seeding cells. After 6 weeks of cell culture, it resulted in a heterogeneous fibrocartilaginous matrix which was similar with the native TMJ disc in structure. Due to the possible effect of remaining PCL scaffold structure, the mechanical properties of the engineered TMJ discs by 6 weeks were approximated to the native properties (Legemate et al., 2016). Schek et al. (2005) used image-based design (IBD) and solid free-form (SFF) fabrication techniques to generate biphasic scaffolds. They found the growth of cartilaginous tissue and bone tissue after seeding different cells which demonstrated the possible therapy to regenerate TMJ joints (Fig. 3). In another study, researchers found that poly (glycerol sebacate) (PGS) might be potential scaffold material for TMJ disc engineering (Hagandora et al., 2013). Considering the complex geometries of TMJ cartilage, AM techniques have great potential in its fabrication, and further exploration is needed in customized TMJ cartilage engineering.

AM application for other craniofacial cartilages: ear, nose and throat.

Other than TMJ, in craniofacial area, cartilage also forms ear, nose, and larynx. Anatomically shaped ear, nose and throat were already printed through PR approaches. PCL-based ear and nose scaffold were printed and perfused with type I collagen containing chondrocytes. The samples were implanted into adult Yorkshire pigs for 8 weeks and histologically analyzed. Histological evidences present that they resulted in the growth and maintenance of cartilage-like tissue (Zopf et al., 2015). A bionic ear was printed with precise anatomic geometry of a human ear by alginate as matrix with 60 million chondrocytes per milliliter. An electrically conductive silver nanoparticle (AgNP) was also printed and infused inductive coil antenna as the sensory part of the ear, connecting to cochlea-shaped electrodes supported on silicone. After in vitro culture, this printed bionic ear not only demonstrated good biocompatibility, but also exhibited enhanced auditory sensing for radio frequency reception, which mimicked the functional human ears (Mannoor et al., 2013). Functional tissue-engineering tracheal reconstruction has also been reported on rabbits by 3D printed PCL scaffolds. The shape and function of reconstructed trachea were restored successfully without any graft rejection. Histological results showed proper cartilage regeneration (Chang et al., 2014).

Technique challenges for cartilage printing and current strategies.

A highlight in cartilage printing is that cells can be printed together with gels as cell vectors. For printing of cell-laden material, the important criterions lay on the suitable shear force and temperature. Otherwise, damage may occur to cells and reduce the viability in the printed constructs (Derby, 2012; Pati et al., 2015). Some studies have been paying attention to modification of the printer nozzle and materials. In one study, an electrospun head was added on an inkjet printer and print electrospun PCL film with fibrin–collagen hydrogel-based cartilage layers inside. It was designed for printing a fibrin-collagen hydrogel of five layers in only 1 mm thickness. With this multi-layer scaffold, this research successfully enhanced the strength of printed materials and overcame the major limitation of inkjet printer in material’s loading ability. Therefore, it is possible to be used to print some load bearing tissue such as cartilage (Xu et al., 2013) (Table 2).

Single dental tissue regeneration.

Lee et al.’s (2014) group has done tooth and periodontal regeneration by cell homing. The research starts from bioprinting of PCL-HA material into two kinds of anatomically tooth shaped scaffold by SLA technology, one is human molar scaffold, and another is rat incisor scaffold. Growth factors of bone morphogenetic protein-7 (BMP7) and stromal cell-derived factor-1 (SDF1) were added into the scaffold to active cell homing in vivo. These two scaffolds were orthotopically and ectopically implanted into mandibular incisor extraction socket and dorsum subcutaneous pouches of rats. After 9 weeks, tooth-like structures and periodontal integration were successfully generated by their study with endogenous cell homing and angiogenesis (Kim et al., 2010). High survival rates were reported in a self-defined shape engineered pulp, which was as high as 87% ± 2%. This research was done to establish a dental pulp like tissue with human dental pulp cells (hDPCs) in sodium alginate/gelatin hydrosol (8:2), and an amount of 1 × 10⁶ cells/ml were seeded (Xue et al., 2013). In a recent study to generate artificial periodontal ligament (PDL) tissue, human PDL cells were seeded on anatomically FDM printing PCL/HA scaffolds. In periodontal osseous fenestration defects on nude mice, guided fiber alignment was later observed oblique orientation to the root surface 6 weeks post implant, which mimics the mature PDL fiber aliment (Park et al., 2014b). Another study invested the osteogenic potential of human dental pulp stem cells (hDPSCs) on different porous PCL printing scaffolds. This research used a specially designed double-layer scaffold system for better osteogenic differentiation. The first layer was nanostructured porous PCL (NSP-PCL) scaffold, and the second layer was PCL coating with a mixture of hyaluronic acid and beta-TCP (HT-PCL) scaffold. With 21 days of in vitro cultivation, the NSP-PCL and HT-PCL scaffolds promoted osteogenic differentiation and Ca²⁺ deposition, showing promising application periodontal tissue regeneration (Jensen et al., 2014). A very recent clinic case first showed the SLS printed PCL scaffolds’ application on a periodontal tissue regeneration in a periodontitis patient. The case demonstrated a 3 mm gain of clinical attachment and partial root coverage. However, the scaffold became exposed at the 13th month and been removed. However, it showed huge potential of AM applications for dental tissues (Rasperini et al., 2015) (Fig. 4).

Combined dental tissue regeneration.

Lee et al. (2014) established a multiphase scaffold mimicking cementum/dentin interface, PDL and alveolar bone by 3D printing blended polycarprolactione/hydroxyapatite (90:10) materials. By adding adequate growth factor and culturing cells, they established PDL-like tissue, the fiber of which connects from one side dentin/cementum tissue to another side bone-like tissue, which is just similar to living PDL’s anatomical property (Lee et al., 2014). Another recent 3D bioprinting research showed BMP7 was benefitional for cementum formation. This research established an interface between cementum and human PDL like tissue, which is novel in combining natural tissue with artificial AM tissue in vitro. The AM scaffold was fabricated with PLGA, and then seeded human PDLSCs. After 6 weeks of culturing, they found that cementum-like layer can be successfully formed in this interface between cementum and human PDL like tissue. They also found that BMP7 helped in cementum matrix protein 1 secretion in vitro, which may be good for cementum tissue establishment (Cho et al., 2016).