3D Printing of Bone Implants and Replacements

Susmita Bose, Samuel Ford Robertson, Amit Bandyopadhyay. American Scientist. Volume 106, Issue 2. March/April 2018.

They say there’s nothing like the real thing, and this adage couldn’t be more true than for an organism’s body. Some animals have evolved to have great capability to recreate their original forms: Starfish can regenerate arms, and worms cut in half may grow into two separate organisms. The axolotl, a salamander relative found in Mexico, is perhaps the champion regenerator, able to regrow limbs, parts of its face, and even its brain and nervous system. Compared with these animals, humans seem to be lacking in the regrowth department, but this lack has led us to develop alternate means of regaining function in the event of a disability.

Prosthetics-from the Greek root prosthesis, meaning addition-were used as early as 1500 BCE to replace lost limbs. These early devices, such as a hook for a hand or a wooden rod for a leg, did allow some recovery of lost functionality. But modern prostheses are a far cry from their carved predecessors. Doctors and research institutions around the world are now using 3D printing techniques to create custom structures that can help to rebuild or remodel different structures across the entire human body. Creating repairs that can interface with bone is a particularly tricky endeavor. Our group specializes in using 3D printing to improve bone replacements. The manufacturing technique is very different from that used to create early devices, and so are the materials; we use biomaterials to create the implants and replacements.

Bodylike Materials

Biomaterials are natural or synthetic materials designed to come into contact with human body tissue and fluid. They are used to replace, assist in the function of, or augment the function of a part of the body. Biomaterials can be metallic, ceramic, or polymeric, but they must meet another important criterion: biocompatibility. In short, you do not want your device to kill the patient. Materials can fail to meet this requirement in several ways. Some materials are just inherently toxic: These include metals such as lead and mercury, nonmetals such as arsenic and bromine, some organic materials, and radioactive elements. Determining which materials are toxic and which are not is not as straightforward as it seems. For example, not all metals are toxic, at least in small amounts. In addition, many different elements of the periodic table are found in normal human biological systems. More plentiful elements include calcium, chlorine, and magnesium, but even trace elements such as copper, iron, selenium, and lithium are still common in the body. Radioactive materials are off-limits as biomaterials, and even though some can be used as effective target-destroying drugs (as in the case of iodine-131 for thyroid treatments), it is important to note that drugs are not considered biomaterials. To complicate things further, some materials are biocompatible initially, but over time they are degraded into constituents that are toxic to the body. This outcome can occur with polymers that degrade into acidic compounds, and with other organics that kill surrounding tissue or accumulate in the organs. Over time, some metal alloys or ceramic materials such as alumina (A1203) can leach harmful ions that damage tissue.

Once materials are found to be biocompatible, they are further categorized as bioinert or bioactive. Why would we ever use bioactive materials in the body? Although bioinert materials may be good at not reacting with the body, the body also wants nothing to do with some foreign object that has suddenly appeared. With any implantation, injury occurs, and the body’s injury response soon follows. With a bioinert material, the final result is fibrous encapsulation, or scarring, that surrounds the implant. In general, it is best to reduce the degree of scarring so the targeted area can go back to functioning normally. Scarring creates a barrier between the host and the implant, which results in incomplete integration.

Bioactive materials seek to provoke a cellular response in the surrounding tissue, which allows normal tissue growth at the surface, thus ensuring that the implant is integrated by what is called bioactive fixation. So should bioinert materials be avoided altogether? Not necessarily.

An implant often consists of a stem that is inserted into a cavity in a bone, and an articulating surface that fits back into the joint and allows for restored motion. In orthopedic implants, the stem is the anchor, and good tissue ingrowth helps to secure it. Conversely, because the articulating surface needs to move, it requires a frictionless surface. Any tissue growth on that surface defeats its intended purpose. The same is true for corrective lenses that come into contact with body fluid in the eye; in that environment a bioactive material would induce a growth reaction from the body that would cloud the lens. Additionally, materials that come into contact with blood, such as heart valves, must be hemocompatible. These valves are bioinert and very smooth, which prevents platelets from attaching to them and red blood cells from breaking on them, which could lead to clotting.

Each category of biomateriał has its own uses, and certain manufacturing methods work best for each material; in addition, different biomaterials need different types of surfaces. Fortunately, there are a number of methods of 3D printing, and the technique can be tailored to the material. The term 3D printing gets thrown around a lot, but what makes it so special? Is it truly better?

Comparative Manufacturing

In traditional manufacturing, metals and thermoplastics are typically formed by first melting them down and then casting, molding, extruding, or shaping them. Because ceramics generally have high melting temperatures, they are extruded or molded as slurries or pastes before they are solidified using a heating and compression process called sintering. Biomaterials also have been processed with these techniques for decades; however, they leave much to be desired.

Although traditional manufacturing methods are well established, in general the dies used for molding, casting, and extruding parts are costly to create, and thus the only way to be cost-effective is to produce a large volume of product. But, should any part’s geometry need to be changed, new manufacturing equipment has to be made, and in some cases the whole process must be redesigned. Additionally, mass production inherently means that all of the parts produced are nearly identical, something that humans are not.

One process that mitigates universality and is used to create devices of complex geometry is computer numerical control (CNC) milling. But this process can be wasteful and expensive, because it works by starting with a block of material and removing whatever is not in the final part.

An example of the limits of traditional manufacturing can be found in titanium, which is used in a wide range of implant applications, from plates and screws to pacemakers and artificial heart pumps. It is a difficult metal to machine and therefore is most often cast or forged, and then ground and polished. But these processes require that the part be fully dense, which means added weight, and that the surface finish be smooth. Natural tissue has a difficult time securing to such implants. So manufacturers must either selectively polish surfaces, or use postmanufacturing modification techniques to enhance tissue-implant integration.

Much of the waste and standardization inherent in the different forms of traditional manufacturing can be addressed by 3D printing. The industrial version of this process is known as additive manufacturing, because it does just that: Implants, devices, and even live tissues are made by adding layers of new material with little-to-no removal of material. Thus complex, one-off designs can be made with very little setup or material waste.

3D printing can also mitigate the weight and density issue. Natural skeletal bone is never completely dense; rather, it has evolved ways of being hollow yet strong enough to support thousands of pounds. Natural structures, bone included, are almost always distinguished by a hierarchical structure, wherein simple building blocks are arranged into increasingly complex levels of design. Traditional manufacturing tries to mimic this structure by adding additional processes during building, but because 3D printing is additive, it has the potential for hierarchical structures to be built into its very core.

The Family Tree of 3D Printing

Most people assume that all 3D printers are those desktop machines now seen everywhere, busily extruding a stream of plastic across a platform. But there are many more types of 3D printers, each with its own method for creating a threedimensional object in real time. They do, however, share some common features. Every print starts with an image, which may be a computer-aided design (CAD) or a 3D image such as a medical magnetic resonance imaging (MRI) or computed tomography (CT) scan, and that image is then converted to a format the printer can read. This new format, commonly a file type called STL (an abbreviation for stereolithography, one of the earlier forms of commercial 3D printing), takes the 3D image and cuts it into a stack of slices. In each slice, an exact path for the printer head is determined that takes into account the limits of the printer’s mobility, while also conforming to the required porosity and strength for the part. Once the file is fully converted, the printer can go on its merry way, rendering the product before one’s eyes. Should the design need to change, a simple modification to the CAD file is all that is needed before new pieces can be printed.

The first mention of 3D printing in terms of the basic principles that are used today was in an article published in 1981 by Hideo Kodama of the Nagoya Municipal Industrial Research Institute in Japan. He described a method of using a tub of photopolymer (plastic resins that can be hardened by light exposure) and a targeted, automated ultraviolet light source to build 3D objects one layer at a time. In 1987, the first printing technology hit the commercial scene: stereolithography. Charles Hull, a prolific inventor, conceived the idea in 1983 while working in a lab that tested photocured coatings for tabletops. His idea, like Kodama’s, was to build up a part layer by layer, but Hull additionally described how these 3D-printed parts could be generated from CAD files, and later he went on to invent the STL file format.

Around the same time came two other printing methods, selective laser sintering and fused deposition modeling. A young undergraduate mechanical engineering student named Carl Deckard at the University of Texas at Austin conceived the idea of Selective laser sintering in 1984. The concept was to partially melt (or sinter) plastic particles together to rapidly create a prototype. Deckard and his faculty mentor, Joseph Beaman, worked together on the design and filed for a patent in 1986; in 1989 the company DTM (short for desktop manufacturing) began producing selective laser sintering machines.

Fused deposition modeling was invented by another student, Scott Crump from Washington State University. He and his wife Lisa formed the company Stratasys, Ltd., in 1989, and their product became the most widely recognizable form of 3D printing. Today’s common desktop 3D printers, such as MakerBot and many others, are based on Stratasys’s technology, a process in which a thermoplastic wire is heated and extruded onto a building platform.

In the years following the advent of these machines, several other methods for manufacturing parts via an additive process have been developed, including selective laser melting, electron beam melting, laser engineered net shaping, bioplotting, and binder-jet printing, among others. The first three methods are similar to selective laser sintering in that a laser or electron beam is used to melt certain metal, polymer, or ceramic powders. Bioplotters are used largely by the biotech and biomedical community to print 3D constructs of soft biological materials. They operate by extruding or dispensing droplets of gel matrix that can contain cells. The last method is the one that our research group primarily focuses on.

Binder-jet printing can use metal, ceramic, or polymer powders. It works by rolling powder from a feed bed over a build platform; then a printhead similar to the one found in an inkjet printer moves over the build, dispensing a binding agent in the exact pattern of the object. After the build is finished, the part must be heated to cure the binder. This turns the printed part into what is referred to as a green part; it can be handled and the excess powder removed. What follows is a binder removal heating phase and final sintering to produce a fully dense part. For certain applications, the use of a powder in this printing method can be more adaptable than use of the gel required by other printers, and the powders also can be ones that are not compatible with laser or electron melting.

Better Bone Scaffolds

In a few years, all of the bone that is in your body today will be gone; it will have been completely recycled and replaced with new bone. Bone is not something passive that serves solely as support for the rest of the body; rather, it is very much alive and responds to its environment by healing, growing, and producing cells.

And it is just as susceptible to injury and disease as any other organ. According to the National Cancer Institute and the American Academy of Orthopaedic Surgeons, in the United States alone there are 53 million people living with or at high risk for osteoporosis; more than 7 million patients have received total hip or knee implants; and more than 100,000 patients are living with bone cancer. These numbers are only liable to rise with an increasing elderly population.

The need for newer, better treatments for chronic diseases affecting bone is the focus of our research. More specifically, we aim to use scaffolds made of novel or natural materials that induce bone repair and coax the tissue into using its natural regenerating mechanisms. These scaffolds act as temporary structures for bone to grow on until the scaffolds are fully absorbed by the body and new bone tissue has taken their place.

Replacing bone is no trivial issue; it is a peculiar biomateriał. Very little of it consists of water, in comparison with the rest of the body, and about two-thirds of bone by weight is inorganic mineral, the majority of which is hydroxyapatite [Caio(P04)6OH2]. This mineral is a member of a larger group called calcium phosphates, each of which contains varying ratios of calcium to phosphate. Calcium phosphates are widely used in bone tissue engineering because of their chemical similarity. In the body, the mineral actually exists as carbonate-substituted hydroxyapatite, in which ions of fluorine, sodium, magnesium, iron, carbonate (C032′), and other substances are substituted into the molecule. The other major components of bone are collagen-which connects to hydroxyapatite crystals to form long, composite strands that give bone both its rigidity and flexibility-and water. And in addition to having a complex chemistry, bone has a unique hierarchical structure (see figure on page 115).

Cells live within the bone and constantly repair and remodel its structure. Osteoblasts are cells that line the surface of the bone matrix. They are responsible for synthesizing the bone matrix both by laying the cross-linked collagen, or the osteoid, and by mineralizing it to form new bone. Bone generation by osteoblasts is balanced by the activity of osteoclasts, the cells whose job it is to remove damaged or old bone so that new bone may take its place. In addition, osteocytes live inside the lamellae, the concentric rings of bone matrix, and help control bone resorption and deposition.

The process that bone undergoes to repair an injury can help to illustrate the full bone remodeling mechanism. As bone ages, it develops more microcracks. Osteocytes detect the stress that these microcracks produce in the material, triggering the release of signaling factors that set into motion the recruitment of osteoclasts. The osteoclasts acidify the local area, causing the bone to be resorbed until signaling factors or hormones such as estrogen initiate their apoptosis (programmed cell death). During the osteoclast recruitment, mesenchymal stem cells are activated and begin differentiating into osteoblasts. After the osteoclasts are terminated, the osteoblasts adhere to the surface and form a tight lining; then they begin to rebuild the bone. (See the figure above.)

The Printing Process

A primary goal when one is trying to mimic bone is to match the chemistry as closely as possible. To this end, the family of calcium phosphates is the first place to look. However, these ceramic materials are inherently a pain to process. Manufacturing is limited to casting with a slurry or paste, and the resulting green part is only loosely held together-but it is at this fragile phase, before sintering, that any machining must be done. Stresses and microcracks are thus reduced during sintering, but such machining often requires the use of more potent binders; it is easy to damage the part irreparably.

A second goal is to make the finished part mimic the geometry of bone. But the complex geometry of natural bone precludes any type of casting method that could be done with ceramic slurries. In this arena, 3D printing particularly outshines its traditional predecessors, because its layer-by-layer building process places nearly no limitations on what shape it can fabricate. In the lab, it takes us about 48 hours to fabricate custom parts out of a ceramic powder-ß-tricalcium phosphate (СазР040Н), which closely resembles the chemistry of natural bone. Without any additional processing steps, these custom parts are ready for their end use: implantation into bone.

Below the glass window on the printer’s lid, a roller spreads fine white powder from the feed bed over the dull metallic plate of the build bed. Over and over, the build bed lowers, anticipating the new layer of powder rolled, right to left, from the rising feed bed. Soon the metal is no longer visible, having been replaced by a smooth surface of flourlike powder. The printhead stirs and glides back and forth over the build bed, from the back of the printer to the front, dispensing droplets of binder-too small to be seen by the naked eye-in the predesigned shape of the first layer. Each printed layer is roughly circular, and hardly visible, only distinguishable by the powder’s slight discoloration from the binder. As it builds, the shape of the part is obscured by the layer of white, unused powder. In just a few hours, the printhead returns to its resting position, and 20 small, off white cylinders only 5 millimeters in length are left in the build bed. When the printhead comes to rest, signaling the end of the process, the build bed is raised. The brick of loosely packed powder is lifted and then transferred to the oven to cure the binder, giving the small cylinders in the center the necessary strength to withstand direct handling. After curing, the block is transferred to a pan, where the loose powder is gently brushed away, revealing the small cylinders bit by bit, in much the same way that an archaeologist dusts earth away from a fossil. The loose powder is collected for later use, and the cylinders are taken to remove the remaining powder.

As gentle air from a compressor whisks away the excess, the cylinder’s true design becomes clearer. Small, square holes arranged in neat rows and columns take shape on all sides. When depowdering has been completed, the part is moved to a furnace for sintering, where the temperature will reach 1,250 degrees Celsius. Once the part cools, it is removed, slightly smaller than when it entered (the consequence of powder particles fusing together); it is now hard and extremely sturdy. Holding it inches away from the eye, one can dearly see through the lattice of pores. The 20 scaffolds are now ready to be implanted, capable of bridging the gap between defects in bone that cannot otherwise heal.

If our goal is to ensure that the chemistry of the implant is as close as possible to that of natural bone, why are we not using hydroxyapatite? It turns out that being too close to nature isn’t ideal if the goal is to have the implant eventually dissolve. In bodily fluid, ß-tricalcium phosphate has a rate of solubility that makes the scaffold degrade and be resorbed by the body at a rate similar to the rate of regeneration of bone. The objective for resorbable scaffolds is to provide support when the bone is weak, and to steadily transfer that load to the bone as it recovers.

A major challenge with ceramic bone scaffolds is that they are brittle and not very tough: Even bearing a small amount of weight can cause them to fracture. This challenge is exacerbated by the necessity of having pores in the scaffolds. If the scaffolds were fully dense, they would be stronger, but bone cells could only live and grow on their outer surface; thus both tissue regeneration and resorption of the implant would take many months. Therefore there is a tradeoff between strength and healing rate. By optimizing the pore size, we have been able to create scaffolds with enough strength to be used in low load-bearing applications such as implants in facial or cranial bones; these scaffolds are fully dissolved by the end of the bone healing process, which lasts 12 to 16 weeks.

The Ion-Gene Connection

Our research also looks at how to tailor the chemistry of the biomateriał, which allows us to control how the cells react to the implant and even to determine what cell types are expressed around the wound. If the balance between the bone-laying osteoblasts and bone-resorbing osteoclasts can be tipped in favor of osteoblasts, then in theory healing time can be reduced. With that in mind, our research has shown that the addition of trace elements, or dopants-such as oxides of silicon, strontium, zinc, and magnesium-to ß-tricalcium phosphate can increase the regeneration rate of bone tissue and the number of new blood vessels formed at the wound site.

These same elements are found in natural bone in trace amounts, and have been shown to play a crucial role in the bone remodeling process in healthy bone tissue. What is difficult to determine is how these ions work inside the body to help enhance bone regeneration. We are working on identifying possible biological pathways and determining whether these dopants work synergistically. By studying the effects that a doped material has on bone cell gene expression, we can tease out the mechanisms by which key gene expression is being increased (upregulated) or decreased (downregulated).

During this work it is important to remember that the body is very good at maintaining homeostasis (equilibrium). So, although increasing the growth rate of osteoblasts is important, eventually those osteoblasts will reach a point where they begin to signal to other osteoblasts to undergo apoptosis, differentiate into osteocytes, and even produce hormones that activate osteoclasts. These mechanisms are all natural ways for the body to make sure it does not overproduce.

What we have shown in our research is that the products of gene expression that are crucial to the development of mature osteoblast cells-for example, products expressed by the gene RUNX2-can be increased by using silicon, zinc, and strontium ions. The ability of these ions to accomplish that is thought to be caused, in part, by the outermost valence electrons of the ions, which are similar in size and charge to the outermost valence electrons of those of calcium ions. Calcium is an important molecule in many different cell-signaling pathways in the body. Therefore, ions of silicon, zinc, and strontiüm are thought to work in conjunction with calcium to continue to positively activate signaling channels that calcium usually mediates. The resulting signaling cascade upregulates two products of gene expression-runt-related transcription factor 2 (RUNX2) and osteoprotegerin (OPG). This upregulation in turn helps preosteoblasts differentiate into mature osteoblasts. The same cascade simultaneously decreases expression of the receptor activator on a protein called nuclear factor kappa beta ligand (RANKL), which reduces the generation of osteoclasts. Ions of magnesium and silicon have been shown to help bone regeneration because they promote the revascularization of tissue by activating vascular endothelial growth factor receptor 2 (VEGFR2). VEGRF2 is important in other signaling pathways that lead to increased production of nitric oxide, a chemical known for its ability to increase local blood flow.

Broadening the Structure

The scope of 3D printing biomaterials extends beyond bone tissue engineering. Research is ongoing regarding the printing of prosthetic limbs, drug delivery devices, artificial bladders, kidneys, cartilage, skin, and trachea. With each application, there are different accompanying problems as well as similar challenges. The printing of living cells requires a controlled environment and incubation, whereas the printing of implants and limbs requires reproducible mechanical properties. For all printed biomaterials, issues such as sterility, printing with multiple materials, and, perhaps most importantly, standardization can be challenging.

Artificial bone is a great example of the challenge of multimaterial printing. Implants so far have not incorporated a material analogous to collagen, which makes bones both tough and resilient to stresses. Orthopedic surgeon and biomedical engineer Michael Yaszemski of the Mayo Clinic in Minnesota and other researchers have taken advantage of printers geared toward printing polymers. They work to combine plastics with bioceramics to produce composite implants that theoretically combine the strength and bioactivity of the ceramics with the flexibility and toughness of polymers.

Another clue to where 3D printing may be heading is the “skin gun” developed by the German biotechnology company RenovaCare. The skin is another complex organ comprising dozens of cell types. A skin gun, which is similar to an airbrush, is used to apply a patient’s own stem cells to a body area damaged by severe burns; initial tests have shown that its use can regenerate skin with minimal scarring within a few days.

Although the skin gun is not a type of 3D printing, it points to the possibility that organs and body parts could be regenerated on site. Some issues that stand in the way include the ability of a printer to keep all materials sterile, and the more challenging problem of how to get U.S. Food and Drug Administration approval for these one-off builds. If the builds for the implants are standardized, how can we ensure that the implanted device will function as intended? That is a question that has researchers in the field debating whether patient-specific implants are even necessary.

But the more mature the field of biomaterials becomes, the closer it gets to using primarily natural constituents. No matter what lengths we go to, the body simply does not like to be in contact with foreign materials. It may be that our bodies were not built to use them; after all, nature has had a 4-billion-year head start on perfecting natural materials. Even now, albeit slowly, technology is heading toward a point at which it will be possible for all the materials we need to be taken from our own bodies, and we will be able to multiply our own tissues to replace themselves. It makes sense; as they say, there’s nothing like the real thing.