Powder-based Additive Manufacturing for Development of Tailor-made Implants for Orthopedic Applications †

Powder-based additive manufacturing (AM) technology has increasingly attracted attention because it realizes fabrication of arbitral three-dimensional structures with a great degree of freedom at high speed, which is almost impossible using conventional manufacturing tools. In early times, AM was known as a rapid prototyping (RP) which merely aimed at a trial prototyping of products mainly from resin materials. Today, thanks to the progress in the heat source, optical system, raw material production, etc., the expected role of AM has dramatically changed into a direct fabrication of final or near-final products made from a variety of materials from resins to ceramics, metals, and intermetallic compounds. Because the powder-based AM is a bottom-up manufacturing method, it offers a big advantage for the small-lot production of a wide range of products; therefore, tailor-made product portfolio is a potent target of AM. In this review paper, the advantage and possibility of AM in the development of functional metallic biomaterials for orthopedic usage is described. Medical treatments need to be flexibly done in accordance with individual pathological conditions and constitutions, therefore, the function of a medical device should be tunable, which is quite a strength of AM.


Introduction
Additive manufacturing (AM) technology could knock the bottom out of the design concepts of the Monozukuri (that denotes manufacturing) in Japan, where it is a specialty.However, despite the fact that the first principles of AM were demonstrated in Japan in the early 1980s (Kodama, 1981), research and technical development in this field is now, ironically, lagging far behind that in other places such as Europe and the U.S. Recognizing this trend, the Japanese government has embarked on a cross-ministry program, which is centered around the Ministry of Economy, Trade and Industry, with the aim of supporting AM technology for materials ranging from resins and gels to ceramics and high-melting-point metals using 3D printers.
AM technology is used to create products of any shape using a near net shape process based on computer aided design (CAD) data, and can be thought of as a system that encompasses all aspects from the design concept and then to the most advanced processing work.As a result, AM is able to provide products that could create new markets, and the new business models and business opportunities.
However, as AM is a moldless technique, it could also greatly change the conventional Monozukuri landscape, and as such there is a concern within certain industries that the use of particular industrial processes could decline when AM transforms the structure of the manufacturing industry.
The emergence of AM technology will change Monozukuri from mass production of a narrow range of products to the small-lot production of a wide range of products.This change is expected to foster the advancement of fields concerned with individuality, including customization and design, and fields that require complex shapes using highly responsive materials.For example, in medical fields involved with creating organs of different shapes and sizes on an individual basis, tailor-made medical care that is capable of achieving the optimal shape and optimal function for each patient is a priority.Additionally, technology that contributes greatly to a wide variety of fields such as food, transport, machinery materials, and tailor-made home electronics is also anticipated.
In this paper we describe the current state of research and development using AM technology pertaining to metal bio-functional materials, focusing on the area of orthopedic surgery.

Types of additive manufacturing
According to ASTM F2792-12a (2009), AM manufacturing is classified into the following seven disciplines: (1) powder bed fusion, (2) binder jetting, (3) directed energy deposition, (4) sheet lamination, (5) vat photopolymerization, (6) material extrusion, and (7) material jetting.When divided according to the material used, AM is classified into areas such as: (1) metal AM, (2) ceramic AM, (3) resin AM, (4) gel AM, (5) composite AM, and (6) bio-3D printers (cell 3D printers).A feature of AM technology is that only the portions required are selectively solidified.One method produces a structure from metal raw powders by irradiating a laser or electron beam as a heat source.One method produces a structure from photocurable resin, via photopolymerization using light containing UV.Another method creates a ceramic or metal product by dispersing ceramic or metal powder in resin, fabricating it into any desired shape, and then removing the resin by heat treatment.These methods can be thought of as new fabrication processes that contrast with cutting work.
In the early AM era, known as rapid prototyping (RP), AM technology did not extend beyond the design of models using resin.Although resin is relatively easy to use, the range of AM applications has now expanded to include ceramics and metals, which have high melting points and are difficult to prototype.Since ceramics and metals can be employed for the actual structural material parts, the value of AM technology has undoubtedly increased through its expanded application with such materials.Bio-3D printers are a type of AM technology that has recently received considerable attention for use in layering cells three-dimensionally.Until now, cells were mostly cultured two-dimensionally, but to exhibit the functions of true organs they need to be layered in three dimensions with intercellular connections and an extracellular matrix between them (Matsusaki et al., 2013).Although it will take time before bio-3D printers are used in vivo as a regenerative medicine technique, it is anticipated that they can be used in ex vivo applications, such as creating mini-organs to assist in drug design and to assess toxicity and efficacy of cosmetics.Most importantly, since it is becoming possible to create blood vessels and arrange multiple cells, the fabrication of artificial organs is now becoming a reality.

Summary of metal additive manufacturing methods (powder bed fusion)
Powder bed fusion, which uses an electron beam or laser beam as a heat source and a metal powder as a starting material, has gained attention as a metal AM technique, and ways of controlling the shape and microstructural parameters of metal materials has been widely studied.
Directed energy deposition is a technique (developed primarily in Europe) that is used as a method for inserting metal powder into a laser beam heat source.Both techniques are capable of fabricating any shape by selectively melting and solidifying metal materials, but the former technique laminates metal powder and then selectively melts and solidifies it, whereas the latter method deposits molten metal at required locations.
Fig. 1 is a schematic diagram illustrating the process used in fabricating a metal product with a lattice structure by powder bed fusion.In this technique, an electron beam (or laser beam) is scanned based on 2D slice data created from 3D CAD data of the final form, and by repeating the process of selectively melting and solidifying one layer of metal powder at a time, multiple layers are stacked up to produce a three-dimensional structure (Murr et al., 2009).Fig. 2 shows raw metal (Ti-6Al-4V) powders utilized for powder bed fusion using an electron (a) and laser (b) beam as a heat source.Generally, comparing to the laser beam, the electron beam generates a larger molten pool because of higher energy density (Guo and Leu, 2013), therefore, a large lamination thickness (~0.1 mm) can be used.Accordingly, raw powder particle used is larger in the case of usage of the electron beam.The surface precision is influenced by the powder particle size.
Three-dimensional metal AM using powder bed fusion is capable of creating high-precision surface shape designs in addition to the structures and porous bodies of complex shapes, which is difficult to achieve using conventional casting and cutting work techniques.Fig. 1(f) shows a metal structure fabricated using the powder bed fusion.This was designed in CAD, and is precisely fabricated with 15-mm side lengths and 3-mm lattice spacing.With metal materials, the control of mechanical functions is possible from the viewpoints of both structure and materials because the hierarchical microstructure can be controlled in addition to the macroscopic shape.

Possible medical applications of additive manufacturing
The major factors involved in the remarkable adoption of AM technology are the generalized use of CAD (due to its ease of use), and the formation of networks capable of transmitting digital data at high-speeds.Three-dimensional design systems that employ 3D CAD modeling can also be used in AM devices at remote sites by converting to an STL format, which is a mode that approximates a 3D freeform surface as a composite of triangular elements.
AM has the following advantages over the production of 3D structures using conventional mechanical process-ing with cutting tools: (1) it can fabricate 3D shapes with free-form surfaces and complex internal structures that are difficult to produce using cutting work; (2) as it is a semi-automated process, specialized knowledge and training are not required to operate the machinery; (3) it is able to produce a desired structure economically and in a short period of time; (4) it does not cause tool wear, generate noise or vibration, or produce shavings; and (5) a high percentage of the materials used can be recycled.
These advantages make it tremendously effective in the customization of medical devices, and there have been numerous attempts by various countries (including Japan) to develop biomaterials using AM technology.For example, 3D bone data obtained from magnetic resonance imaging (MRI) or computed tomography (CT) can be used to design an implant shape in CAD that fits the bones of a patient, and to actually produce such a device using AM technology.Novel attempts at using such technology have been examined for bone replacement devices such as artificial hip joints including the acetabular cup and head (Marin et al., 2010), total knee replacement implants (Murr et al., 2011), and for skull bones and jawbones (Parthasarathy et al., 2011;Jardini et al., 2014;Nickels, 2012).Another strength of AM technology, as described below, is the ability to hierarchically control (Zhang et al., 2013) not only the exterior shape but also the internal structure, including the porous structure.
Although concerns have been voiced about cited drawbacks, such as the small amounts of voids remaining in compacted portions, and the degradation of fatigue char-acteristics, compaction can be achieved by the optimization of fabrication conditions together with the use of hot isostatic pressing (HIP) after fabricating (Ye et al., 2010), and mechanical properties exceeding those of conventional cast materials can be obtained.

Creation of artificial hip joint fit to medullary cavity
The greatest advantage of metal AM technology is that it can provide tailor-made implants that conform to the bone shape of individual patients.Because most implants currently produced are designed based on an average bone shape, they do not necessarily fit everyone, and therefore for most patients the bone needs to be cut to match the shape of the implant.Although it is possible to use precision processing to customize the implant shape for a patient, it is rarely worth the cost.In contrast, metal AM enables fabricating using a near net shape process based on 3D CAD data, even for complex shapes with many free-form surfaces, and is extremely effective as a manufacturing process of products that conform to the bone shape of individual patients.
Fig. 3 shows an artificial hip joint fabricated by electron beam melting (EBM) and customized to conform to the femoral medullary cavity.Both the exterior shape and a porous structure enabling bone ingrowth can also be formed simultaneously on the surface at specified locations.Guidelines on AM implants are currently being created under the guidance of the Japanese Ministry of Health, Labour and Welfare (NIHS, 2013) , and it is considered that implants customized by AM will be commercially available in the near future.

Creation of biomimetic hierarchical structure using powder metal
When powder bed fusion is used, the layered metal powder is partially melted and solidified.Therefore, with the exception of completely solid structures, a residue of the starting metal powder is generated inside the structure.Normally the powder is removed and only the solid part is extracted as a product, but as the structure becomes more complex and compacted, and as the proportion of the periphery covered with the solid part becomes greater, the more difficult it becomes to extract the unmelted powder.With AM, however, it is possible to modify the preconceived notion of removing unnecessary metal powder and using only the structural framework, and to actually utilize the powder that would previously have been discarded (Nakano et al., 2011a).This conceptual transformation can be used to make it possible to achieve a hierarchical structure and to impart novel mechanical functionality.
Hierarchical structures are seen throughout the natural world, and these hierarchical organizations and structures organically unite and function on various scales.Bone tissue and plants are both typical examples of hierarchical structures (Bar-On and Wagner, 2013).With EBM, it is possible to design a structure that contains a small amount of starting powder at an order of tens of microns.The fabricated beam (or wall structures) formed by melting this powder are then several hundred microns in size, and the exterior shape measures several mm to several cm, or greater.As a result, it is possible to construct a hierarchical structure in which structural elements on different scales work together both structurally and functionally.
For example, as shown in Fig. 4(a), a product was created by neck formation between powders that had a low Young's modulus and an energy absorbency similar to bone.This was achieved by confining metal powder with a particle volume fraction of approximately 50 % in the structure, and then performing appropriate heat treatment.In addition, in an attempt to make a bone mimetic structure, a structure with uni-directional pores was designed to exhibit the uniaxial anisotropy of the mechanical function possessed by long bones (Bonfield and Grynpas, 1977).
Fig. 4(b) shows stress-strain curves of the products in which an unmelted powder is packed fabricated by EBM.The product without heat treatment (as-EBM) exhibits a low energy absorptive capacity and low Young's modulus.However, the heat treatment below the α/β phase transformation temperature promotes the necking of unmelted residual powder, which results in the great improvement of energy absorptive capacity evidenced by the emergence of a plateau in the stress-strain curve.
The powder part of as-EBM material behaves like space and does not affect mechanical properties.However, the necking achieved between powders enables control of the Young's modulus and the ability to impart an energy absorptive capacity.The deformation behavior of the structure differs greatly depending on the neck-size ratio.In this case, the threshold value of plateau emergence is approximately 0.25.However, if the neck-size ratio is increased to approximately 0.5, compaction of the powder part progresses, resulting in the disappearance of the plateau.Neck-size ratio is governed by volume diffusion, and the heat treatment temperature is its controlling factor.Therefore, the mechanical function of a structure depends on the construction of a network between powders that occurs with an increased heat treatment temperature.Bone mimetic structures fabricated in this manner, with low Young's moduli and high energy absorptive capacities, are called powder/solid (P/S) composites, and this method is expected to become the new way of using powder bed fusion (Nakano et al., 2011a).

Creation of a bone replacement porous structure with a low Youngʼs modulus
The appeal of using a titanium alloy as a biomaterial is related to its relatively low Young's modulus (which is close to that of living bone), high biocompatibility through osteointegration (by which it adheres directly to bone on a level seen by optical microscope), and its osteoconductivity.Bone replacement materials using titanium alloys are widely used, and a variety of products that are designed using EBM are now beginning to use clinically (Unwin, 2014).
Titanium alloys have a lower Young's modulus than general metal materials, but still exhibit a higher value than the 20 to 30 GPa of long bones.For example, an alloy of Ti-6Al-4V has a Young's modulus of 110 GPa.As a result, implanting this alloy in bone produces stress shielding, which hinders normal stress transfer to osteocytes (OCYs) as mechanosensors in bone, thereby leading to reduced bone mass, reduced bone mineral density (BMD), and reduced bone quality around the implant (Noyama et al., 2012).
Reducing the Young's modulus of a structure by introducing pores is an important technique (Ikeo et al., 2014).To succeed in reducing the Young's modulus with the goal of reducing stress shielding, control of mechanical function focused on pore formation into the structure using various techniques has been attempted.However, since nearly all conventional methods aim at making porous structure using a self-assembly reaction, it is difficult to impart a certain designed pore diameter or porosity, let alone desired pore configuration.In addition, although pore formation by self-assembly can control the internal structure to a certain extent, it is unable to customize the shape.
However, using AM technology it is possible to freely control the macroscopic exterior shape using a 3D shape design on a PC.Additionally, simultaneous control of the exterior shape and the internal hierarchical structure is possible by controlling the irradiation conditions of the heat source beam.Based on this concept, 3D porous structures with any desired pore diameter and porosity can be fabricated using layer-by-layer additive manufacturing (Ikeo et al., 2012).Fig. 5 is a photograph of a structure with uni-directional pores fabricated by powder bed fusion using an electron beam as a heat source.By varying the scan interval of the electron beam, a structure can be produced in which the lattice spacing is controlled.Porosity can be continuously varied, even in cell structures and the lattice structures of more complex 3D shapes (Murr et al., 2012).In such structures, the Young's modulus can be reduced by the required amount by reducing the cross-sectional area ratio while taking fatigue characteristics into consideration.In addition, the bending strength can be maintained by designing a shape with consideration of the second moment of inertia with respect to a bending moment.Furthermore, by matching the anisotropy of the bone microstructure (collagen/apatite) (Nakano et al., 2002), it is also possible to control osteoconductivity.

Creation of osteoconductive porous metal implant
In producing a porous metal implant, it is possible to artificially fabricate a porous structure at an order of several hundred microns and above.Pores of the smallest formable size promote entry of osteoblasts into the structure, and therefore production of a new bone matrix.Pores on a larger scale function as scaffolds for osteoblasts, enabling free design and fabrication of the exterior shape, and thus may allow for the creation of tailor-made implants suited to the bone shapes and bone defect sizes of individual patients.
Fig. 6 illustrates the appearance of a porous cylindrical metal implant made of Ti-6Al-4V by EBM (powder bed fusion) after being implanted for 24 weeks in a rabbit ulna defect site (Nakano et al., 2011b).The implant was tailormade according to a CAD design, and has a crosssectional diameter of 8 mm ϕ and various lengths of 10 mm, 15 mm, and 20 mm (according to the varying bone defect sizes) (Fig. 6(a)).Fabrication of this porous metal cylindrical implant is not possible without using a 3D AM.Its design criteria were: (1) to have an apparent Young's modulus similar to that of cortical bone, (2) the ability to function as a scaffold for constructing the preferential orientation of collagen/apatite, so that osteoblasts are arranged and migrate preferentially along the long axis of the bone, (3) creation of a structure that stimulates the formation of the medullary cavity for maintaining hematopoietic function; and (4) creation of a bone implant with a tailor-made exterior shape that conforms to the bone defect size.
As shown in Figs.6(b) and (c), the new bone formation and bone fusion is dependent on whether or not the porous cylindrical metal implant is inserted.The bone matrix is seen to exhibit a uni-directional orientation of the collagen/apatite crystals along the long axis of the bone, resembling normal bone.Additionally, formation of a medullary cavity is also seen inside the porous body, and the porous implant is utilized without hindering bone functions such as sustaining mesenchymal stem cells and maintaining hematopoietic function (Fig. 6(e)).Because this metal porous implant uses a titanium alloy powder with a relatively large particle size (averaging approximately 80 μm as shown in Fig. 2(a)) and is formed by EBM, it has a relatively coarse surface.However, it has also been reported that such a coarse surface structure leads to promotion of calcification and osteointegration (Alla et al., 2011), and as such this is a major factor that needs controlling in AM.

Imparting osteoconductivity by controlling implant surface shape
Designing the interface between living tissue and an implant is a critical factor in extracting the maximal functionality of living tissue.The bone/implant interface is no exception.A porous coating or a roughened surface is often introduced onto the medial part of an artificial hip joint to provide an anchor effect.However, bone mass, BMD, and the bone quality including collagen/apatite crystal orientation varies depending on the OCYs' stress-sensing (Nakano et al., 2012a).For this reason, when constructing an implant surface intended to provide just a simple anchor effect, loosening sometimes occurs at the interface.Here, note that bone quality as referred to in this study is related to factors that determine bone strength other than BMD and bone mass (NIH consensus 2000(NIH consensus , 2001)).
In short, the surface shape of an implant is critical for early fixation after implantation, and the long-term main- tenance of bone quality.Developing an implant that actively promotes stress transfer to the bone, rather than producing stress shielding also enables promotion of sound bone microstructure, in addition to avoiding the functional degradation of the bone after implantation (Noyama et al., 2013).Introduction of oriented pores and grooves on the implant surface along the principal stress direction which provides a space that delivers the optimal environment for bone cells, could serve as an effective means, since the collagen/apatite orientation of bone (which is an index of bone quality) is constructed parallel to the maximum principal stress vector (Nakano et al., 2002).It is of note that in many cementless artificial hip joints, although the porous coatings are introduced onto the proximal portions, stress shielding and subsequent degradation of bone mechanical function is inevitable (Noyama et al., 2012) because the conventional porous coatings consist of random pores that do not achieve ideal stress distribution to promote stress transfer and formation of new bone with oriented microstructure similar to healthy bone.Fig. 7 shows a titanium alloy artificial hip joint for a beagle dog, into which oriented grooves were introduced using cutting work or AM.When this type of hip joint implant with oriented grooves was implanted for 24 weeks (cutting work) or 4 weeks (AM) in the beagle femur, the greatest amount of new bone was seen inside the oriented grooves running along the principal stress vector (60°) (Noyama et al., 2013).The distribution of apatite crystal orientation in the oriented grooves (as analyzed using microbeam X-ray diffraction with a transmission optical system) is expressed as a radar diagram, which shows that its preferential orientation runs along the oriented groove walls parallel to the principal stress vector.This means that apatite/collagen creates a uni-directional orientation along the oriented grooves, and therefore the long-term fixation of the implant is expected.
Even if cytokines are used to stimulate bone regeneration, the recovery of collagen/apatite orientation as an index of bone quality is very slow with respect to BMD recovery, and therefore a long time is required for recovery of the bone's mechanical function (Nakano et al., 2012b;Ishimoto et al., 2013).Osteoblasts can be aligned and an oriented bone matrix resembling intact bone can be produced using anisotropic steps on the metal surface (Matsugaki et al., 2012), or by application of uni-directional stress in vitro (Matsugaki et al., 2013), which is governed through a certain mechanisms in which cells, molecules, and genes are involved in.Improving the bone matrix surrounding the implant by leveraging the bone orientation mechanism is very useful for enabling long-term fixation of implants.
Metal AM also makes it possible to fabricate complex surface shapes, and construction of the preferential orien-tation of apatite in bone has been seen where oriented grooves were formed in the direction of the principal stress vector.Formation of the oriented bone matrix parallel to the principal stress vector (wall direction of oriented grooves) has been seen, similar to the oriented grooves formed by cutting work.However, although surface roughness is greater than when using cutting work, oriented grooves on the order of several hundred microns can be introduced by direct metal fabrication using EBM (Fig. 7(d)).In short, a complex implant surface shape designed in consideration of the stress response function of living bone can be realized.It is known that the form and microstructure of living bone is primarily controlled in the direction of loading through bone's functional adaptation capacity (Tsubota et al., 2009).Therefore, it is important in the development of bio-functional metal implant devices, to optimize their external shape and internal microstructure for aiming at controlling in vivo principal stress distribution and inducing the healthy bone tissue possessing the anisotropic microstructure.AM technology utilizing powders is one of the most promising tools for making medical implant devices that are morphologically and functionally tunable based on the individual patient's conditions.

Conclusion
Because individuals have unique bone shapes and bone matrix characteristics, customization of implants is necessary for tailor-made medical care.With AM technology, the realization of ideas for achieving bone functionalization is possible by designing surface shapes, reducing the Young's modulus of the implant (by making a powder/ solid composite for limiting stress shielding), as well as guiding cell migration to make the bone matrix sound.Using a combination of cell printing and regenerative medicine it is expected that the importance of such customization will be increasingly realized.Additionally, AM technology that enables the small-lot production of a wide range of products can also be applied to transport machinery materials, home electronics, and so forth.In an era that calls for de-individuation, AM technology capable of fabricating individual custom products holds infinite possibilities, and can also create a potential impetus for new market developments in the pursuit of high added value.
social system reform and research and development, Support Program to break the bottlenecks at R&D System for accelerating the practical use of Health Research Outcome, from the Japan Science and Technology Agency (JST).

Fig. 1
Fig. 1 Schematic diagram illustrating the powder bed fusion method used in fabricating a metallic product with a lattice structure.Reproduced from Ikeo et al. (2014) with modifications.

Fig. 2
Fig. 2 Starting raw metal (Ti-6Al-4V) powders utilized for powder bed fusion using (a) an electron and (b) laser beam as a heat source.

Fig. 3
Fig.3A model of a tailor-made artificial hip joint fabricated by electron beam melting (EBM).The exterior shape of the stem is designed to fit to marrow cavity.A porous structure on the stem surface enabling bone ingrowth can be fabricated by one-step EBM process.

Fig.
Fig. (a) Schematic illustration of an EBM fabricated product in which an unmelted powder is remained and photograph of the powder part after heat-treated.(b) Nominal stress-nominal strain curves of the products with and without heat treatment that makes powders necked.The product with the necked powder inside becomes highly energy absorptive due to the stress transmission between powders through the necks.Reproduced from Nakano et al. (2011a) with modifications.

Fig. 5
Fig. 5 Photograph of porous products fabricated by EBM with various-sized uni-directional pores.

Fig. 7
Fig. 7 (a) An artificial hip joint (stem) for beagle with oriented grooves on the proximal surface and (b) simulated principal stress in the grooves by finite element analysis (FEA).The bone with oriented apatite in the groove depth direction is formed inside the groove fabricated by (c) machining and (d) EBM method.Reproduced from Noyama et al. (2013) with modifications.