WO2009151604A1

WO2009151604A1 - Methods for making ceramic articles, including ceramic scaffolds for bone repair

Info

Publication number: WO2009151604A1
Application number: PCT/US2009/003501
Authority: WO
Inventors: Yunzhi Yang; Yongxing Liu
Original assignee: Yunzhi Yang; Yongxing Liu
Priority date: 2008-06-12
Filing date: 2009-06-10
Publication date: 2009-12-17

Abstract

Ceramic porous and non-porous articles are made by a process that includes a hardening step in which the liquid of a liquid-containing ceramic composition is extracted from the ceramic composition by exposing the ceramic composition to a solvent in which the liquid in the composition is soluble before the ceramic composition is solidified into the final ceramic article. Calcium phosphate porous ceramic articles of the invention are constructed and characterized that may be used as implants to repair bone defects.

Description

METHODS FOR MAKING CERAMIC ARTICLES, INCLUDING CERAMIC SCAFFOLDS FOR BONE REPAIR

Field of the Invention The present invention pertains to the field of fabricating ceramic articles and particularly to the field of fabricating porous ceramic articles which may be used for various purposes, such as a scaffolding for many different applications, such as for tissue engineering and bone replacement and repair. In one particular embodiment, the invention pertains to the field of biodegradable ceramic scaffolds, such as calcium phosphate based scaffolds, that are useful in the treatment of skeletal defects.

Background of the Invention

Ceramics are used extensively in a large number of industrial applications. They are used as building materials, as cements and mortars, as abrasives, and in recent years ceramics have been developed for specialized uses in such fields as electronics, communications, and medicine.

In medicine, biodegradable macroporous ceramic scaffolds have been used as engineered grafts for tissue engineering, particularly bone tissue engineering. Such scaffolds typically are made with hydroxyapatite (HA) or tricalcium phosphate (TCP) , or a combination of HA and TCP, with additives such as silica, magnesium, sodium, potassium, and zinc. The porous nature of these scaffolds permits the ingrowth of vascular and structural tissues and, because the scaffolds are biodegradable, can be used safely and without the need to remove the implant from the body.

For bone repair, particularly for defects in the spine and long bones, such as the bones of the legs, it is critically important that a ceramic scaffold implant have a high compressive strength and that this strength is maintained as the implant is biodegraded before the bone itself has healed and has sufficient strength. However, there is an inverse relationship between porosity and mechanical strength of the implants as the mechanical strength decreases as the porosity and pore size increases. In addition, biodegradable synthetic bone implants decrease in strength as the implant is degraded by contact with body fluids. Loss of strength of an implant at a time before the healed bone is able to support weight or support itself can lead to failure of the implant and of the repair process. Ma, U.S. Patent No. 6,673,285 discloses a method for fabrication of porous articles, such as polymer scaffolds. Ma discloses that the scaffolds may be made by casting a composition onto a negative replica of a desired macroporous architecture of the porous article to form a body, and that the negative replica, referred to as a porogen, is removed, thereby forming the porous article. Ma discloses that this method may be utilized to form a porous article from various materials, including polymers, ceramics, glass, and inorganic compounds . The inventors have utilized the method of Ma in order to attempt to make macroporous ceramic calcium phosphate (CaP) scaffolds. Such attempts, however, were unsuccessful and this process could not be used to form a sintered integrated ceramic body. It was found that the ceramic article produced in this manner lacked sufficient hardness and strength, and broke into a multiplicity of pieces before and during sintering .

Various scientific articles describe methods of manufacture of macroporous ceramic (CaP) scaffolds of various porosity and report on the compressive strength of these scaffolds. See, Hing, J. Mater. Sci . Mater. Med., 10(3) :135- 145 (1999); Liu, Ceramics International, 23:135-139 (1997); Seplveda, J. Biomed. Mater. Res., 50:27-34 (2000); Ramay, Biomaterials, 24:3293-3302 2003); Almirall, Biomaterials, 25:3671-3680 (2004); Cyster, Biomaterials, 26:697-702 (2005); Silva, Biomaterials, 27:5909-5917 (2006); Uemura, Biomaterials, 24:2277-2286 (2003); Sous, Biomaterials, 19:2147-2153 (1998); Guo, Tissue Engineering, 10:1830-1840 (2004); Kwon, J. Am Ceramic Soc, 85:3129-3131 2002); and Milosevski, Ceramics International, 25:693-696 (1999) . These reports show that the strength of porous CaP scaffolds tends to decrease with increasing porosity and that most of the scaffolds produced by the prior art methods have a compressive strength of only about 0.8 to 8 MPa (megapascals) with one report of a scaffold having 70% porosity, pores not completely interconnected, and a compressive strength of about 11 MPa.

Large bone defects that result from disease or damage can be replaced or reconstructed by a structural graft or prosthesis. Use of a patient's own bone as the source of a graft, referred to as an autograft, remains the "gold standard" of graft choice due to its excellent osteogenicity, osteoinductivity, and osteoconductivity . However, the use of autografts is limited in clinical situations by the lack of available bone for harvest, particularly in the case of children and large-scale defects, significant postoperative morbidity at donor sites, increased operative time and blood loss, and additional cost. An alternative to autografts is the use of bone from another individual, referred to as an allograft. However, the preparation of an allograft requires donor screening, sterile harvesting, and processing, and presents an increased risk of infections and disease transmission, as well as inconsistency in quality. As a result of these problems, biomimetic synthetic bone grafts are desirable . Calcium phosphate (CaP) ceramics are attractive alternatives for artificial bone scaffold construction. CaP is the main inorganic component of vertebrate calcified hard tissues. The CaP materials used most frequently in clinical settings are beta-tricalcium phosphate (TCP) , hydroxyapatite (HA) and their composites. The degradation of CaP by dissolution does not produce any known harmful effects. Sterilization and shelf storage of the materials do not present difficulties and there is no risk of disease transmission or of an immunogenic response. Additionally, CaP scaffolds can be used to deliver living cells and growth factors to the implantation site.

It is of critical importance that the CaP scaffold has a macroporous structure to permit bone growth into and onto the scaffold. Conventional techniques for fabricating 3- dimensional CaP scaffolds include foaming, sacrificial templates, replication of polymer foams by infiltration with CaP slurries, hydrothermal conversion of either coral or bone, and replamineform. However, the resulting porous structures are typically rather random in architectures with regards to pore sizes, shapes, alignment, and interconnectivity .

Robocasting, a solid freeform fabrication technique, has been developed to fabricate HA scaffolds and show potential for better controlling pore size, shape and a customized fabrication. However, this method requires expensive 3D freeform manufacturing systems and special CaP ceramic slurries for the machine. Consequently, this method has not been widely adopted.

A significant need remains for a method for producing a CaP scaffold for bone repair applications that provides control over the architecture and composition of the scaffold and that can be used to provide a scaffold that mimics the physical and chemical properties of bone.

Brief Description of the Drawings Figure 1 is a photograph showing, on the left side, a ceramic composition slurry in a plastic tube container prior to drying, in the middle, a green body dried by the solvent extraction step of a method of the present invention, and on the right side, a green body dried by exposure to air at room temperature without the solvent extraction step of the invention .

Figure 2 is a graph comparing the compressive strength in MPa of a porous ceramic article made by a method of the invention with porous ceramic articles made by methods other than that of the invention. The arrows point to data points for the porous ceramic articles made by a method of the invention .

Figure 3A is a 3-dimensional computer-reconstructed Micro CT (computed tomography) image of a dense scaffold showing the lack of pores made to mimic the structure of cortical bone. Figure 3B is a top view 2-dimensional Micro CT image of the dense scaffold. Figure 3C is a side view 2-dimensional Micro CT image of the dense scaffold. Figure 3D is a 3-dimensional computer-reconstructed Micro CT image of a two-zone graded ceramic scaffold having pores in the inner zone and lacking pores in the outer zone made to mimic the structure of bone. Figure 3E is a top view 2-dimensional Micro CT image of the graded scaffold. Figure 3F is a side view 2-dimensional Micro CT image of the graded scaffold.

Figure 4A is a 3-dimensional computer-reconstructed Micro CT image of a porous scaffold with pores of 600 μm to 800 μm. Figure 4B is a top view 2-dimensional Micro CT image of the porous scaffold. Figure 4C is a side view 2-dimensional Micro CT image of the porous scaffold. Figure 4D is a 3-dimensional computer-reconstructed Micro CT image of a porous scaffold with pores of 350 μm to 500 μm. Figure 4E is a top view 2- dimensional Micro CT image of the porous scaffold. Figure 4F is a side view 2-dimensional Micro CT image of the porous scaffold.

Figure 5 is a scanning electron microscopy photograph showing the solid struts and interconnectivity between pores of a scaffold made by a negative replica method. The arrows indicate the solid struts. Figure 6A is a 3-dimensional computer-reconstructed Micro CT image of a radially graded porous ceramic article in which an inner zone of the article contains pores between 350 μm to 500 μm in diameter and an outer zone contains pores between 600 μm and 800 μm. Figure 6B is a top view 2-dimensional

Micro CT image of this radially graded porous ceramic article. Figure 6C is a corresponding Micro CT side image.

Figure 7A is a 3-dimensional computer-reconstructed Micro CT image of a radially graded porous ceramic article in which an inner zone of the article contains pores between 600 μm and 800 μm in diameter and an outer zone contains pores between 350 μm to 500 μm in diameter. Figure 7B is a 2-dimensional Micro CT top to bottom image of this radially graded porous ceramic article. Figure 1C is a corresponding 2-dimensional Micro CT side image.

Figure 8A is a 2-dimensional Micro CT side image of a vertically graded macroporous ceramic article in which the top portion has smaller pores of 300 μm to 400 μm and the bottom portion has larger pores of 600 μm to 700 μm. Figure 8B is a top view 3-dimensional computer-reconstructed Micro CT image of the vertically graded macroporous article showing the smaller pores at the top surface. Figure 8C is a bottom view 3-dimensional computer-reconstructed Micro CT image of the article showing the larger pores at the bottom surface. Figure 9A is a scanning electron microscopy photograph of a compositionally graded porous ceramic article. Figure 9B is a graph that indicates the varying composition of the article at various numbered locations as shown in Figure 9A.

Figure 10 is a graph showing the dissolution behaviors of porous TCP scaffolds following immersion in Tris buffer for 4 weeks. A shows the dissolution behavior of the scaffolds with uniform 600-800 μm pores. B shows the dissolution behavior of the scaffolds with uniform 350-500 μm pores. C shows the dissolution behavior of the graded scaffolds with central 350- 500 μm pores and peripheral 600-800 μm pores. D shows the dissolution behavior of the graded scaffolds with central 600- 800 μm pores and peripheral 350-500 μm pores.

Figure 11 is a series of photographs showing the morphological changes of graded CaP scaffolds that occurred in vitro. Cl is a graded scaffold with central 350-500 μm pores and peripheral 600-800 μm pores. C2 is the scaffold of Cl following immersion in acidic buffer medium at pH3. Dl is a graded scaffold with central 600-800 μm pores and peripheral 350-500 μm pores. D2 is the scaffold of Dl following immersion the acidic buffer medium.

Figure 12 is a non-decalcified histological examination of CaP scaffolds showing morphology changes that occur following subcutaneous implantation of the scaffolds for a period of one month. A shows results observed for the scaffold with uniform large pores of 600-800 μm. B shows results observed for the scaffold with uniform small pores of 350-500 μm. C shows results observed for the scaffold with graded pores having central small pores of 350-500 μm and peripheral large pores of 600-800 μm. D shows results observed for the scaffold with graded pores having central larges pores of 600-800 μm and peripheral small pores of 350- 500 μm.

Figure 13 is a graph showing the initial loading of BMP-2 onto scaffolds of different pore sizes. * indicates significant differences (P<0.05) .

Figure 14 is a graph showing the cumulative elution of BMP-2 from scaffolds of different pore sizes.

Figure 15 shows BMP-2 induced ectopic bone formation in non-decalcified porous CaP scaffolds at one month after implantation. Al, Bl, Cl and Dl are micro CT images; A2, B2, C2 and D2 are histology pictures obtained with Anderson' s rapid bone stain counterstained with acid fuchsin. Al and A2 are of a scaffold with uniform 600-800 μm large pores. Bl and B2 are of a scaffold with uniform 350-500 μm pores. Cl and C2 are of a graded scaffold with central 350-500 μm pores and peripheral 600-800 μm pores. Dl and D2 are of a graded scaffold with central 600-800 μm pores and peripheral 350-500 μm pores.

Figure 16 shows radiographs taken at 2 weeks (A) and 4 weeks (B) after implantation of a CaP scaffold constructed by a method of the invention into a defect in the radius. Healing of the radial defect is apparent after two weeks and after four weeks.

Figure 17 shows micro CT images of healing of a defect of the radius following implantation with a CaP scaffold constructed by a method of the invention. A represents a cross-sectional view. B represents a longitudinal view.

Description of the Invention In one embodiment, the present invention is based upon the discovery by the inventors that removing a portion or all of the liquid present in a fluid ceramic composition by extraction with a solvent having a lower surface tension than the liquid, thereby obtaining a hardened ceramic composition, followed by solidifying the resultant hardened ceramic composition, such as by the application of heat, results in a ceramic article possessing unexpectedly higher strength than that possessed by similar ceramic articles that are made without the solvent-based liquid removal step. The presently disclosed method permits the manufacture of strong ceramic articles for use in various industries, such as for building construction, electronics, telecommunications, and in the manufacture of housewares. The presently disclosed methods are especially useful in the manufacture of biodegradable materials such as for implantation into the body, such as for porous implants such as those used for bone reconstruction and regeneration techniques. In one embodiment, the invention is a method for making a ceramic article. According to this embodiment of the invention, a fluid ceramic composition containing a liquid is formed into a desired shape and is exposed to a solvent in which the liquid of the ceramic composition is soluble at a concentration and for a time sufficient to extract some, a majority of, or all of the liquid from the composition. Typically, but not necessarily, the liquid from the composition is replaced by an equal volume of the solvent. Following the extraction, the resultant "dried" composition, is caused or permitted to solidify to form a ceramic article.

In another embodiment, the invention is a ceramic article made by the method of the invention. Such a ceramic article may be referred to herein as "solvent-hardened", which term indicates that, prior to solidifying to form the ceramic article, the fluid ceramic composition that was used to make the article was exposed to a solvent in which liquid in the composition was soluble at a concentration and for a time sufficient to extract the liquid from the composition and, following this extraction, the composition was caused or permitted to solidify to form the solvent-hardened ceramic article .

In another embodiment, the invention is a method for making a macroporous CaP scaffold having high interconnectivity and mechanical strength. According to this embodiment, a CaP scaffold is made by the negative replica method disclosed above by using a negative replica that is defined by a template comprising a multiplicity of discrete porogen particles. Preferably, as disclosed above, a hardening step utilizing an extraction solvent is performed prior to the final curing of the scaffold, typically prior to removal of the negative template from the ceramic composition. An illustrative preferred method for making the macroporous CaP scaffold is described in Example 1 below. In another embodiment, the invention is a macroporous scaffold, such as a CaP scaffold, made by a negative replica method, which method includes a hardening step utilizing an extraction solvent that is performed prior to the final curing of the scaffold.

In another embodiment, the invention is a method for treating a skeletal defect. According to this embodiment of the invention, a macroporous CaP scaffold of the invention is implanted into or onto a bone within the body of an animal in need thereof and the implanted scaffold is permitted to remain in place in or on the bone for a time sufficient for new bone made by the animal to develop on the scaffold.

As used herein, the term "removing" when referring to a liquid of a fluid ceramic composition refers to reducing the concentration of the liquid in the ceramic composition. The removing may be accompanied by a replacement of the volume of liquid removed with a smaller, equivalent, or higher volume of another liquid.

As used herein, the term "extract" when referring to the liquid of a fluid ceramic composition, means to reduce the concentration of the liquid in the ceramic composition by exposing the ceramic composition containing the liquid to a solvent in which the liquid of the ceramic composition is soluble. Such extraction is preferably performed by immersing a container containing the ceramic composition into a larger container containing the solvent. This method of extraction typically, but not necessarily, results in a dilution of the concentration of the liquid in the ceramic composition by providing a larger volume into which the liquid will dissolve. Generally, but not necessarily, the volume of the liquid that is removed from the ceramic composition will be replaced by the solvent. The extraction may also be performed by any other method by which a liquid may be extracted by use of a solvent in which the liquid is soluble. Examples include pouring the solvent into the container containing the ceramic composition, or by spraying.

As used herein, the term "ceramic material" refers to an inorganic non-metallic crystalline or partly crystalline, or glass, material that either solidifies upon cooling from a molten mass or that forms a solid structure due to the action of heat. There are innumerable examples of ceramic materials, all of which are intended to be within the scope of the present invention. Examples of ceramic materials include aluminum silicates, zirconium oxides such as zirconium dioxide, aluminum oxides, titanium oxides, tantalum oxides, carbides, borides, nitrites, and suicides, calcium ceramics such as calcium nitrite, calcium sulfate, calcium hydrogen sulfate, calcium hydroxide, calcium carbonates, calcium hydrogen carbonate, and calcium phosphates, alkali metal hydroxides, alkaline earth hydroxides, disodium hydrogen phosphate, disodium hydrogen phosphate dodecahydrate, disodium hydrogen phosphate heptahydrate, sodium phosphate dodecahydrate, dipotassium hydrogen phosphate, potassium phosphate tribasic, diammonium hydrogen phosphate, ammonium phosphate trihydrate, sodium bicarbonate, barium titanate, bismuth strontium calcium copper oxide, boron carbide, boron nitride, ferrite, lead zirconate titanate, magnesium diboride, silicon carbide, silicon nitride, steatite, uranium oxide, yttrium barium copper oxide, and zinc oxide.

As used herein, the term "ceramic article" refers to an article of manufacture that is made from a ceramic material. A ceramic article has a glazed or unglazed body of crystalline or partly crystalline structure, or of glass, which body is produced from essentially inorganic non-metallic substances and is either formed from a molten mass that solidifies upon cooling or is formed and simultaneously or subsequently matured by the action of heat.

As used herein the term "ceramic composition" refers to a composition comprising a ceramic material that flows sufficiently for casting purposes. The ceramic composition may be a solution or a non-solution and may be, for example, in the form of a melt, a slurry, or a flowable paste, which may be made by wetting a powder of a ceramic material with a liquid. The ceramic composition may contain additional components, such as binders, plasticizers, anti-flocculants, and lubricants.

The liquid of the fluid ceramic composition may be any liquid or multiplicity of liquids into which a ceramic material may be dispersed, with or without the use of additional materials such as a binder, plasticizer, anti- flocculant, or lubricant. Preferably, the ceramic composition includes a binder, which is typically a polymer, which may be water miscible or immiscible, and which may be hydrophilic, hydrophobic, or amphiphilic. Examples of water soluble binders include polyvinylpyrrolidones (PVP), polyvinylpyrrolidone/vinyl acetate copolymers, polyvinyl alcohols (PVA), carboxymethyl celluloses, hydroxypropyl cellulose starches, polyethylene oxides (PEO) , polyacrylamides, polyacrylic acids, cellulose ether polymers, polyethyl oxazolines, esters of polyethylene oxide, esters of polyethylene oxide and polypropylene oxide copolymers, urethanes of polyethylene oxide, and urethanes of polyethylene oxide and polypropylene oxide copolymers. A preferred binder is carboxymethyl cellulose (CMC) . Additional examples of suitable polymer binders, which may or may not be water soluble, include one or more of polypropylene (PP) , amorphous polypropylene (APP), polyolefin (PL), polyethylene (PE), ethylene vinyl acetate (EVA), polystyrene (PS), polyvinyl acetate (PA), polyvinyl alcohol (PVA), polyphenylene oxide (PPO), methyl cellulose (MC), hydroxyethyl cellulose (HEC), polyacrylate, apolyacrylamide, poly (lactide-co-glycolide) (PLGA), poly (lactide) (PLA), polyglycolic acid (PGA), polyanhydrides, poly(ortho ethers), polycarprolactone, polyethylene glycol (PEG) , polyurethane, polyacrylic acid, polyethylene glycol, polymethacrylic acid (PMMA), alginates, collagens, gelatins, hyaluronic acid, polyamides, polyvinylidene fluoride, polybutylene, and polyacrylonittrile . The liquid of the fluid ceramic composition may be water miscible or immiscible and may be one or more organic or inorganic solvents or solutes. The fluid composition may contain a multiplicity of liquids. The liquid may be an aqueous liquid. For example, the liquid may be water or may be a combination of water and organic or inorganic acids or alcohols. Examples of polar organic solvents and solutes that are suitable for the liquid of the fluid ceramic composition include alcohols such as methanol, ethanol, propanol, isopropanol, and butanol, carboxyl acids, sulfonic acids, compounds containing an -OH, -SH, enol, or phenol group, formic acid, 1,4-Dioxane, tetrahydrofuran, acetone, acetonitrile, dimethylformamide, and dimethyl sulfoxide. Examples of non-polar organic solvents and solutes include hexane, benzene, toluene, diethyl ether, chloroform, ethyl acetate, and dichloromethane . Examples of inorganic solutes are hydrobromic acid, hydrochloric acid hydroiodic acid, nitric acid, sulfuric acid, perchloric acid, boric acid, carbonic acid, chloric acid, hydrofluoric acid, phosphoric acid, pyrophosphoric acid, ammonium hydroxide, alkali metal hydroxide, alkaline earth hydroxide, disodium hydrogen phosphate, ammonia, methylamine, pyridine, disodium hydrogen phosphate, disodium hydrogen phosphate dodecahydrate, disodium hydrogen phosphate heptahydrate, sodium phosphate dodecahydrate, dipotassium hydrogen phosphate, potassium phosphate tribasic, diammonium hydrogen phosphate, ammonium phosphate, trihydrate, sodium bicarbonate, NaHCO₃, NaHS,

NaHSO₄, NaH₂PO₄, Na₄HPO₄, NH₄OH, NH₄H₂PO₄, (NH₄) ₂HPO₄, NH₄HCO₃, and NH₄HSO₄.

The fluid ceramic composition is formed into a desired shape by any method by which the desired shape may be formed. The desired shape may be any three-dimensional form. In order to make this form, the composition may be rolled, pulled, pressed, or molded to form a shape such as wire. The ceramic composition may be formed on a relatively planar surface or within a liquid, or may be cast upon an irregular non-planar template.

In many applications, it is desirable to obtain a porous ceramic article. Such products are useful for electrodes and supports for batteries and solid oxide fuel cells, for scaffolds for bone replacement and tissue engineering, for heating elements, for chemical sensors, for solar radiation conversion, and for filters in the steel industry. Porous ceramics may be made by replica methods, using either a positive replica or a negative replica of the ceramic article. With the positive replica technique, a porous template, such as a sponge, is coated with a fluid ceramic composition. The ceramic composition may or may not contain additives such as binders and plasticizers that provide strength and flexibility to the coating so that it will not crack during subsequent phases of the fabrication process. Following the coating step, the coated sponge is passed through rollers to remove the excess ceramic composition and to form a thin ceramic coating over the struts of the sponge. The ceramic coated sponge is then dried and pyrolysed by heating, typically between 300° and 800° C, which removes fluid from the ceramic composition, removes the replica template from the ceramic composition, and solidifies the ceramic composition. Finally, if desired, the remaining ceramic coating may be densified by sintering at temperatures ranging from 1100° to 1700° C depending on the nature of the ceramic material. The positive replica technique has a disadvantage for certain indications because the struts of a ceramic article made with this technique are necessarily hollow. This results because the ceramic composition coats portions of the template that define the struts. When the template is removed, this leaves a hollow ceramic strut overlying the space where the replica strut previously existed. Also, due to the removal of the porogen strut during pyrolysis, the ceramic struts often crack during this phase of manufacture, which markedly degrades the strength of the porous ceramic article. The negative replica technique does not share these disadvantages. In this technique, a sacrificial porogen is utilized to make a template of the pores of a ceramic article, rather than of the product itself. According to this method, a negative replica of a desired porous ceramic article is made, typically by forming an assemblage of a multiplicity of discrete porogen elements, and casting a ceramic composition onto the assemblage and thereby obtaining a biphasic composition of a continuous matrix of the ceramic composition and a sacrificial phase within the matrix. The sacrificial phase may be distributed homogeneously throughout the ceramic matrix or may be assembled into a defined structure.

Following the formation of the biphasic composition, the matrix ceramic phase must be partially consolidated to form what is referred to as a "green body" or a "body" so that the porous structure of the ceramic composition does not collapse when the sacrificial porogen material is removed. Present methods of consolidation involve the use of setting agents or binders or the formation of a stiff attractive network of particles distributed throughout the matrix. Other methods include the use of sol-gel transitions based on the condensation of metal alkoxide and hydroxides in solution or by a curing process at a temperature slightly lower than that which will melt and remove the porogen materials.

The porogen materials are removed by a means that is selected based upon the nature of the porogen. Organic porogens, such as waxes, are often extracted by pyrolysis by applying long thermal treatments at temperatures between 200° and 600° C. Other sacrificial porogens, such as salts, ceramics, or metallic particles, are usually extracted by chemical leaching. Following the removal of the porogen, the ceramic is typically further processed, such as by kiln-firing or sintering.

Unlike the positive replica method, the negative replica method results in the formation of a ceramic article having struts that are solid, rather than hollow. Therefore, the negative replica method produces porous templates that typically have a higher compressive strength than do ceramic articles of similar porosity formed by the positive replica method. Another advantage of the negative replica method is that it provides precise control over the architecture of the ceramic articles and can be used to produce products that are graded, either functionally or structurally. For example, gradations of pore size within a ceramic article may be obtained by grading the distribution of porogen particles of various sizes within the negative replica. In addition, gradations of composition with a ceramic article may be obtained by grading the distribution of ceramic slurry within the negative replica. In both the positive and negative replica method, the template may be made of any material upon which a ceramic composition may be cast and which can be removed by a method that does not destroy the structure of the resulting ceramic article. Positive templates are typically made of a polymeric sponge, such as polyurethane. Other positive template materials include carbon foam and natural templates such as coral and wood. Negative template porogens include polymers such as poly (lactide) or poly (lactide-co-glycolide) , salts, sugars, and waxes such as paraffin. The presently disclosed methods are applicable to any method for forming a ceramic article, including methods as indicated above in which no template is used and those in which a template is used. If a template is used in the formation of a ceramic article, the presently disclosed methods are applicable to both positive and negative replica template methods.

According to the presently disclosed methods, a hardening step is performed prior to the final curing step of a ceramic article. With non-template methods of forming a ceramic article, such as when making an essentially non-porous ceramic article, the hardening step is performed before the ceramic composition has solidified and while it is still pliable. With template methods of forming a ceramic article, the hardening step is preferably performed prior to removal of the positive or negative template from the ceramic composition. Thus, with negative template methods, the hardening step is preferably performed during the formation of the green body. Because it is desirable that the ceramic composition should be as hard as possible before the template is removed, so as to minimize the occurrence of cracks in the composition, it is not preferred, although it is possible and is within the scope of the present invention, to perform the hardening step of the invention after the template has been removed from the ceramic composition.

In accordance with the methods of the present disclosure, the hardening step is performed by exposing the ceramic composition to a liquid extraction solvent in which non-fluid components of the ceramic composition are practically insoluble or are insoluble and in which the liquid component of the ceramic composition is miscible for a time sufficient to extract the liquid from the ceramic composition. The extraction solvent may, but does not necessarily, replace the volume of the liquid that is extracted from the ceramic composition. If the ceramic composition contains a binder, it is preferred that the binder is less soluble in the extraction solvent than it is in the liquid of the ceramic composition. In a preferred embodiment, the binder is insoluble in the extraction solvent. The amount of time in which the ceramic composition is exposed to the liquid extraction solvent may be varied, depending on several factors, including the materials comprising the ceramic composition, the fluid component of the ceramic composition, the liquid extraction solvent employed, and the degree of hardening that is desired. Preferably, but not necessarily, the hardening step is performed for a time sufficient that the ceramic composition will be sufficiently rigid to maintain its structural integrity in the absence of external support, for example as shown in Figure 1. In the situation where a ceramic composition is combined with a template, the material composing the template is preferably, but not necessarily, practically insoluble or insoluble in the solvent so as not to remove the support of the template from the ceramic composition before the ceramic composition has hardened. If the template material is soluble to some extent in the solvent, then the amount of time that the template is exposed to the solvent should be adjusted so that the strength of the template is not reduced by dissolution to an extent that the ceramic composition is no longer sufficiently supported.

The selection of the particular extraction solvent employed will depend on the identities of the liquid contained within the ceramic composition and of the composition of the template, if present. For example, if the ceramic composition fluid is an aqueous fluid such as water, preferably containing a binder such as carboxymethyl cellulose (CMC) , and the template is composed of paraffin, a preferred extraction solvent is a short-chain alkyl or aryl alcohol, such as methanol, ethanol, isopropanol, butanol, or phenol, or a mixture thereof. For another example, if the ceramic composition fluid is acetone, preferably containing a binder such as polymethyl methacrylate (PMMA) , and the template is composed of sugar or salt, a suitable extraction solvent may be one or more of tetrahydrofuran (THF) , hexane, benzene, or toluene.

Although not wishing to be bound by theory, it is postulated that the hardening of the ceramic composition that results due to the solvent extraction step of the presently disclosed methods relates to the difference in surface tension between the original liquid in the ceramic composition and the extraction solvent. For example, in the case where the original liquid in a fluid ceramic composition is aqueous, water has a relatively high surface tension compared to organic solvents, for example hexane, acetone, or alcohols such as ethanol . When a ceramic composition containing water is dried, the water exerts a force on itself and on solid components of the ceramic composition, creating stress and a tendency for the ceramic composition to crack as water is forced out by evaporation or upon heating. In contrast, replacement of water from the ceramic composition with a solvent having a lower surface tension, such as with an organic solvent, for example ethanol, acetone, or hexane, reduces the cohesive and adhesive forces of the fluid ceramic composition and results in a hardened ceramic composition with reduced stress and tendency to crack. Accordingly, when selecting an extraction solvent, it is preferred that the extraction ■ solvent have a surface tension less than that of the original liquid of the ceramic composition.

The relative surface tensions of liquids of ceramic compositions and extraction solvents may be obtained by reference to published values for surface tensions of liquids. Alternatively, a suitable extraction solvent may be selected based on a test that reflects differences in surface tension of liquids. According to this test, equal volumes of a ceramic material are mixed in separate containers with equal volumes of two liquids, for example water and ethanol to obtain a pourable, viscous slurry. The liquid having the higher surface tension will produce a more viscous slurry that that produced with the liquid having the lower surface tension.

Another characteristic of a preferred extraction solvent is that it should be miscible in the liquid of the ceramic composition. It is also preferred that, if a binder is present in the ceramic composition, such binder should be more soluble in the liquid of the ceramic composition than in the extraction solvent. It is theorized that, when an extraction solvent is used in which the binder is less soluble than the binder is in the ceramic composition liquid, the binder will come out of solution and will function as a glue between particles of the ceramic composition and will contribute to the strength and rigidity of the ceramic composition. Thus, for example, in the case of an aqueous fluid as the liquid of a ceramic composition containing CMC in solution, extraction of water with ethanol results in increased concentration of the CMC in the liquid or a precipitation of the CMC, which causes adherence of particles of the ceramic composition.

The ceramic composition, and the template if present, are exposed to, and are preferably immersed in, the extraction solvent at a temperature below the melting point of the template. Because paraffin typically melts between 47° and 64° C, it is preferred that, if paraffin is the material of which the template is composed, the temperature of the extraction solvent should be less than 50°, more preferably less than 47°, and most preferably less than 45° C.

The concentration of the extraction solvent should be that which is sufficient to cause removal via extraction of the liquid of the ceramic composition. In the situation where the ceramic composition liquid is water, a preferred concentration of ethanol is 70%. This concentration of ethanol has been found to extract water from a ceramic concentration sufficiently to increase the hardness and strength of the resulting ceramic article. If desired, a higher concentration of ethanol may be used, but care should be utilized to ensure that the ceramic composition fluid is not removed so rapidly to crack or deform or otherwise result in structural weakness of the ceramic article.

In one especially preferred embodiment, the liquid in the ceramic composition, with or without an associated positive or negative template, is extracted by exposing the composing to sequentially higher concentrations of the extraction solvent. The stepwise increase in extraction solvent concentration is preferred because a high concentration of the solvent may be utilized in this fashion which more efficiently dissolves fluid from the ceramic composition but does not dissolve the fluid as rapidly as if the ceramic composition had been exposed immediately to the higher concentration of solvent. Thus, the graded drying reduces the potential stress on the ceramic composition due to an overly rapid drying process.

For example, if the extraction solvent is ethanol, the ceramic composition, with or without an associated template, may first be exposed to the ethanol at a concentration of 70%. The ceramic composition may then be removed from the ethanol and then exposed to ethanol at a concentration of 80%.

Alternatively, 95% ethanol could be added to the ethanol that the ceramic composition is in so as to raise the concentration to 80%. Following the extraction with 80% ethanol, further extraction may be performed with 90% ethanol and/or with 95% ethanol. Similar extraction procedures may be used with other combinations of ceramic composition fluid and extraction solvent .

If desired, the extraction fluid may also be utilized to remove a template, such as a sacrificial porogen utilized as a negative replica. By immersing a ceramic composition and replica template in an extraction fluid at a temperature higher than the melting point of the material of which the template is composed, the template will liquefy and will flow out of the ceramic composition and into the extraction fluid. For example, with paraffin as a template, ethanol or other alcohol may be used at a temperature above the melting point of paraffin, which is typically 50°C or higher.

It is preferred that the extraction fluid utilized be one in which the material of the replica template is not soluble. In this way, the extraction fluid and the liquified template will remain in separate phases and can readily be separated from each other. This will allow for easy collection of the template material from the extraction fluid which will allow for both the extraction fluid and the replica template material to be recycled and reused. Removal of the template material in this manner also obviates the need for pyrolysis, burning out the porogen at very high temperatures, which may potentially cause structural defects such as microcracks and therefore reduce the mechanical strength of the ceramic article.

In a preferred embodiment, the extraction of fluid from the ceramic composition is performed utilizing a solvent in which a template material is not soluble at a temperature below that of the melting point of the template material and then the temperature of the extraction fluid is elevated to that above the melting point of the template material during continued fluid extraction. In this way, strengthening of the ceramic composition and removal of the template is performed in a single process. For example, if a paraffin positive or negative replica template is utilized in the fabrication of a ceramic article, the ceramic composition associated with the template may be exposed to 70% ethyl alcohol at a temperature below the melting point of paraffin. This temperature is maintained for a sufficient time to ensure that, when the template is removed, the ceramic composition will be sufficiently strong not to collapse if the paraffin were to be removed. The temperature of the ethyl alcohol may then be increased to a temperature above the melting temperature of the paraffin, which will cause the paraffin to melt. The ethyl alcohol and paraffin may be removed and replaced with successive treatments of higher concentration ethyl alcohol for further extraction of fluid from the ceramic composition, which is now a green body. A composition of the invention is a solvent-hardened ceramic article, that is the article was made by a process in which a liquid-containing ceramic composition is formed into a desired shape and is hardened by exposure to a solvent in which the liquid contained in the ceramic composition is soluble at a concentration and for a time sufficient to extract the liquid from the composition and that following the extraction, the "dried" composition, which is preferably, but not necessarily completely free of liquid, is caused or permitted to solidify to form the ceramic article. The ceramic article made by the method of the invention may be non-porous or porous. If porous, it may be made by any method by which a porous ceramic article may be made so long as the ceramic composition is subjected to the solvent extraction step prior to the final solidification of the composition to form the ceramic article. The porous ceramic article may be made with any desired degree of porosity, from 1% to over 90%. For example for calcium phosphate, as well as other ceramic articles, the porosity may be between 60% and 95%, preferably between 70% and 90%. The porous ceramic articles of the invention may be made to have any desired degree of interconnectivity between pores, up to 100% interconnectivity . The porous ceramic article may be made by a negative replica method in which discrete porogen particles are used to define a template upon which a ceramic composition is cast. One advantage of the negative replica method is that the interconnectivity of the pores may be controlled by heating or otherwise causing individual elements of the sacrificial porogen to coalesce to a desired degree which will correspond to the degree of interconnectivity of pores in the final ceramic article. Another advantage of the negative replica method is that a solvent-hardened ceramic article of the invention may be a porous article having uniformity of distribution of pores, pore sizes, and composition or any of these characteristics of the article may be varied to provide a porous article that varies spatially in the distribution of pores, of pore sizes, and/or of composition. Non-porous articles of the invention may also be compositionally graded.

The ceramic articles produced by the methods of the disclosure have many uses. The increased compressive strength of the ceramic articles of the invention are of use in many fields, including for structural materials for buildings and electronics, as well as for making biodegradable ceramic articles for implantation into the body of humans and other animals . One particular use of the ceramic articles of the invention is for the implantation in order to repair bone. Synthetic biodegradable ceramic bone graft materials made by presently available methods of manufacture have compressive strength less than that of bone. Additionally, the ceramic bone graft materials lose a significant portion of their initial strength over time as the synthetic bone is absorbed into the body. The methods of the present disclosure, utilized for strengthening biodegradable ceramic bone grafts, therefore will provide a significant contribution to this field.

The method of the invention is useful in the creation of macroporous structures which have a high degree of interconnectivity between pores and a high compressive strength. The method of the invention has been utilized in making a sintered macroporous CaP ceramic article by a negative replica method, which articles may have 100% interconnectivity between pores, a porosity up to or even higher than 70%, and solid struts between pores. The inventors have found that similar articles produced by prior art negative replica methods lacking the solvent extraction step of the present disclosure were not sufficiently strong to withstand sintering temperatures. In fact, to the knowledge of the inventor, no macroporous article made by negative replica methods and having 100% interconnectivity between pores has been produced prior to the present invention.

The negative template-casting method of the present disclosure provides for fine control of macroporous structures by varying the sizes of beads utilized and their arrangement. For instance, scaffolds with two ranges of pore sizes, 600-800 μm and 350-500 μm, were successfully fabricated. High interconnectivity of pores was also readily achieved in these scaffolds regardless of pore size. Analysis using scanning electron microscopy (SEM) revealed reticular structure of the scaffolds in which each and every macropore interconnects to multiple neighboring pores. These interconnective windows were at the macroscale, averaging 330 +/- 50 and 440 +/- 57 μm, respectively, dependant on the sizes of paraffin beads. Table 1 describes the physical characteristics of two scaffolds of different porosity fabricated by otherwise identical negative template-casting method of the present disclosure .

* Porosity was estimated by dividing the apparent density by theoretical density of β-TCP (3.156 g/cm³)

Table 1

Various porosities of scaffolds, such as but not limited to between 70% or lower to 90% or higher, can readily be obtained by controlling the template process which is determined by paraffin bead size and arrangement. In addition to macroporosity, microporous structures on struts were also achieved by template-casting method, which may potentially improve the scaffold performance in vivo.

In making one embodiment of the macroporous scaffold of the invention, a multiplicity of particles, such as beads, are arranged to form a negative replica. Typically, but not necessarily, the particles are arranged within a container, such as a tube. The particles are caused to agglomerate, such as by heating the particles to a temperature at which they begin to melt and become tacky, causing adjacent particles to adhere to each other, and thereby forming a unitary mold structure. A ceramic composition, such as a CaP ceramic composition, is then introduced into the container to fill the spaces not occupied by the negative replica.

The porosity of the scaffold may be controlled in various ways. Because the template is a negative replica, the use of larger size particles will provide a template of greater porosity than will be obtained using particles of smaller size. Additionally, increased melting of the particles, such as by increasing the temperature and/or time of heating, will result in increased surface of adherence of one particle to another, thereby resulting in increased porosity.

If desired, a multiplicity of containers may be situated one within another so as to form a multiplicity of zones. Within the different zones, particles of different sizes or shapes may be utilized in order to vary the architecture, such as the porosity, of the mold structure within each zone. Within the different zones, different ceramic compositions may be introduced so as to vary the composition of the scaffold from zone to zone. After the ceramic composition is introduced into the container, the ceramic compositions are exposed to a solvent, as described above, to harden the ceramic compositions and remove liquid that is contained within the compositions. The negative replica is removed, such as by chemical or heat treatment, and the scaffold is permitted to solidify, such as by air drying or sintering.

If desired, the scaffold may be loaded with cells, such as mesenchymal or other stem cells, or with a growth factor, such as bone morphogenic protein (BMP) or an angiogenic growth factor such as vascular endothelial growth factor (VEGF) or transforming growth factor (TGF) . The scaffold may also be loaded with a pharmaceutically active agent, such as an antibiotic or an analgesic.

If desired, the scaffold may be coated, such as with chitosan or other polymer, which coating may facilitate the incorporation of cells, drugs, or growth factor onto the scaffold. If the scaffold is to be coated, the coating is typically applied before loading the scaffold with the cells, drugs, or growth factors. Coating and/or loading the scaffold may be accomplished by any means that provide for coating or loading CaP scaffolds. For example, coating and loading may be spraying, painting, or pipetting the coating material or a liquid containing the loaded material onto the scaffold, or by immersing the scaffold in such a liquid. The immersion method is preferred because the inventor has found that this method provides for more precise regulation of loading and elution based on pore size. The CaP scaffold of the invention may have zones of different architectures, which can be used to control biodegradation, spatial and or temporal, of the implanted scaffold. This permits a temporally and spatially controlled osteogenesis . In a preferred embodiment, the architecture of the scaffold is arranged to form a biomimetic scaffold that resembles the architecture of bone. According to this embodiment, a macroporous scaffold is made having a multiplicity of zones, such as an inner zone and an outer zone. The inner zone has a higher porosity than that of the outer zone. In this way, the inner zone mimics the architecture of cancellous bone and the outer zone mimics the architecture of cortical bone.

The CaP scaffold may be used to repair bone defects. For repair of bone defects, the scaffold may or may not be loaded with a growth factor, such as BMP. The CaP scaffold has been utilized in long bones of a rabbit. Repair of bone defects in the rabbit was obtained utilizing BMP loaded and unloaded CaP scaffolds of the invention. Repair was more rapid, however, with scaffolds that were loaded with BMP.

To further illustrate the above embodiments of the invention, the following examples are provided. It is to be understood that these examples are provided for illustrative purposes and are not to be construed as limiting the scope of the invention.

Example 1 Paraffin beads were prepared by a conventional water- suspension method. The paraffin beads were sifted in order to obtain beads with diameters ranging from 1.2 to 1.8 mm. The sifted beads were filled into polyethylene cylinder tubes. The filled tubes were placed into warm water at a temperature of about 50° C to allow the beads to soften and to coalesce into a unitary mold structure.

A fine tricalcium phosphate (TCP) powder was mixed with distilled water at various weight ratios of 1: (0.2-10) . This mixture was stirred and carboxymethyl- cellulose (CMC) was added at various weight ratios of 1: (20-1) . The mixtures were stirred until a homogenous slurry was obtained.

The slurry was poured onto the top of the paraffin mold. The mold with the slurry was placed into a vacuum chamber for at least 10 minutes, at which time the chamber was filled with air and the paraffin mold was checked to determine if it had been completely filled with the slurry. If not completely filled, additional repetitions of the pouring of the slurry onto the mold and the exposure to the vacuum were performed until it was determined that the paraffin mold was completely cast with the slurry to make porous ceramic bodies for making macroporous ceramic articles.

Another set of samples was prepared by directly filling the slurry into the polyethylene cylinder tubes, without prior filling of the tubes with paraffin beads. These molds, which were not cast upon a negative template, in order to make solid (non-porous) ceramic bodies for making nonporous ceramic articles .

The ceramic bodies, porous and non-porous, were soaked in 70% ethyl alcohol at a temperature between 30° and 60° C for at least 30 minutes. The temperature was then increased to between 60° and less than 100° C and maintained for no less than 30 minutes in order to melt and remove the paraffin molds. The alcohol and melted paraffin were replaced with 80% to 95% ethyl alcohol at 60° to less than 100° C and maintained for at least 30 minutes. The ethyl alcohol was replaced with new ethyl alcohol at the same concentration and maintained for at least 30 minutes.

A control group for each of the solid and porous ceramic bodies was air dried, without applying this solvent-based solidifying and drying fluid extraction process.

All samples were then placed into an electric furnace and were heated to a temperature of 1100° to 1300° C for a period of 3 hours to produce sintered porous and non-porous ceramic articles .

Example 2 - Testing of the ceramic articles of Example 1

The porosity of the porous ceramic scaffolds of Example 1 was calculated by dividing the apparent density of the scaffold with the TCP theoretical density of 3.156 g/cm³ and was determined to be about 73%. The apparent density of the scaffolds were determined by measuring the mass of the scaffold and dividing by the volume of the scaffold. Macromorphology and three-dimensional structure of the scaffolds were determined by micro computed tomography (micro CT) . Scanning electron microscopy was used to determine the microstructure of the scaffolds. Maximum compressive strength of the ceramic articles prepared in Example 1 was determined by using a mechanical tester (Instron 4465, Instron Corp., Canton, MA) . The maximum compressive strength was measured and, for a macroporous scaffold made with the solvent extraction step, having 100% connectivity and having pore sizes of 350-500 μm or 600-800 μm, was determined to be 17 +/- 4 MPa. It was not possible to determine the compressive strength of the similar macroporous scaffold made without the solvent extraction step, because these scaffolds invariably cracked into pieces prior to or during the exposure to sintering temperatures. Figure 1 shows, on the left side, a plastic tube filled with a slurry of a ceramic composition prior to drying, in the middle, a macroporous green body dried by the solvent-extraction method of the invention, and on the right side, a green body dried by exposure to air at room temperature. As shown in the middle of Figure 1, the solvent extraction drying step maintained the integrity of the green body whereas, as shown in the right side of Figure 1, air drying did not maintain the integrity of the green body, which crumbled and cracked into a multiplicity of pieces during construction.

A plastic tube filled with a slurry of a ceramic composition prior to drying is shown on the left side of Figure 1. In the middle of Figure 1 is shown a macroporous green body dried by the solvent-extraction method of the invention and on the right side of Figure 1, a green body dried by exposure to air at room temperature. As shown in the middle of Figure 1, the solvent extraction drying step maintained the integrity of the green body whereas, as shown in the right side of Figure 1, air drying did not maintain the integrity of the green body, which crumbled and cracked into a multiplicity of pieces.

Similarly, maximum compressive strength of a dense non- porous article made with the solvent extraction process of Example 1 was determined to be 297.8 +/- 73.0 MPa. The comparable dense non-porous articles made without the hardening step of the invention invariably developed cracks during sintering and so were not tested for compressive strength.

These results demonstrate that both porous or non-porous ceramic articles (scaffolds) may be made by the method of the present disclosure and that such ceramic articles are able to withstand processes such as sintering. Moreover, they show that articles made by the method of the present disclosure have a very high compressive strength.

Example 3 - Comparison of Strength of Macroporous Scaffolds The compressive strength of additional macroporous CaP scaffolds made according to the method of Example 1 and having a porosity of 73% was tested by the method of Example 2 and determined to be 16.86 MPa +/- 3.60 MPa. This was compared to the strength of prior art macroporous scaffolds made with various methods as reported in the scientific literature. See, [1] Hing, J. Mater. Sci. Mater. Med., 10 ( 3) : 135-145 (1999); [2] Liu, Ceramics International, 23:135-139 (1997); [3] Seplveda, J. Biomed. Mater. Res., 50:27-34 (2000); [4] Ramay, Biomaterials, 24:3293-3302 (2003); [5] Almirall, Biomaterials, 25:3671-3680 2004); [6] Cyster, Biomaterials, 26:697-702 (2005); [7] Silva, Biomaterials, 27:5909-5917 (2006); [8] Uemura, Biomaterials, 24:2277-2286 (2003); [9] Sous, Biomaterials, 19:2147-2153 (1998) ; [10] Guo, Tissue Engineering, 10:1830-1840 (2004); [11] Kwon, J. Am Ceramic Soc, 85:3129-3131 (2002); and [12] Milosevski, Ceramics

International, 25:693-696 (1999) . The results are shown in Figure 2, which is a graph plotting compressive strength in MPa on the Y-axis and porosity in volume % on the X-axis.

As shown in Figure 2, the maximum compressive strength of the macroporous scaffold made according to the method of the present disclosure (indicated by the arrow) is markedly higher than is that of scaffolds constructed using different methods described by others in the art. This is true even when the scaffolds made according to the methods of others had a lower porosity which, because of higher mass per volume, would have been expected to be stronger than higher porosity scaffolds constructed using the methods of the present disclosure. Example 4 - Compressive Strength of Cortical Bone and Biomimetic CaP Scaffold

A dense CaP ceramic article, referred to in this example as a scaffold even though the article lacks pores, was made according to Example 1. Figure 3A-C shows a 3-dimensional and two 2-dimensional Micro CT images of dense scaffold showing the lack of pores. This pore-less scaffold was made to mimic the structure of cortical bone.

A graded CaP ceramic scaffold, containing an outer zone of dense pore-less ceramic and an inner zone of a porous scaffold, was made according to the method described in Example 1. Figure 3D-F is a 3-dimensional and two 2- dimensional MicroCT images of the scaffold showing the two- zone graded ceramic scaffold having 600 μm to 800 μm pores in the inner zone and lacking pores in the outer zone. This two- zone scaffold was made to mimic naturally occurring bone having an inner zone of cancellous bone and an outer zone of cortical bone. The two-zone graded ceramic scaffold was made by filling a tube with paraffin beads followed by filling of the tube with a ceramic slurry and filling an outer concentric tube with the slurry without first filling this outer tube with the beads.

The maximum compressive strength of the dense ceramic scaffold and the two-zone ceramic scaffold was determined as described in Example 2 and was compared to the strength of cortical bone reported in An YH and Draughn, RA, "Mechanical Testing of Bone and the Bone-Implant Interface", CRC Press, Boca Raton, FL (2000) .

The strength of cortical bone reported in An and Draughn is 200 +/- 36 MPa (from 133 to 295 MPa) . The strength of the non-porous dense CaP scaffold was determined to be 297.8 +/- 73.0 MPa. The strength of the two-zone scaffold, mimicking the structure of bone having both cortical and cancellous zones, was determined to be 153.9 +/- 29.2 MPa. The results of this study were surprising because, not only was the compressive strength of the dense scaffold substantially higher than that of cortical bone, the two-zone scaffold also had a compressive strength similar to or somewhat higher than that of cortical bone. It is to be noted that the compressive strength of bone having both cortical and cancellous portions will naturally be less than that of cortical bone alone.

Therefore, the data establish that the CaP scaffold made by the method of the invention has a strength that is equal to or higher than that of bone. This indicates that the scaffolds of the invention should be able to withstand functional loading when used as implants for long bone grafting .

Example 5 - Manufacture of Macroporous Scaffold

Macroporous scaffolds were made according to Example 1 to produce scaffolds having pores between 600 μm to 800 μm, shown in Figure 4A-C, and between 350 μm and 500 μm, shown in Figure 4D-F.

Example 6 - Interconnection of Pores of Macroporous Scaffold

A macroporous scaffold having pores between 600 μm to 800 μm was made according to Example 1 and was imaged by scanning electron microscopy, as shown in Figure 5. The interconnective pore size was determined to be 440 +/- 57 μm. The struts between pores are solid due to formation of the scaffold by the negative replica method.

Example 7 - Manufacture of Radially Graded Macroporous Scaffold

Macroporous scaffolds were made according to Example 1 except that two concentric polyethylene tubes were utilized and paraffin beads of two different sizes were respectively filled into each of the tubes. Figure 6A-C shows a 3-D and two 2-D Micro CT images of a radially graded porous ceramic article scaffold) in which an inner zone of the article contains pores between 350 μm to 500 μm in diameter and an outer zone contains pores between 600 μm and 800 μm. Figure 7A-C shows a 3-D and two 2-D Micro CT images of a radially graded porous ceramic article in which an inner zone of the article contains pores between 600 μm and 800 μm in diameter and an outer zone contains pores between 350 μm to 500 μm in diameter .

Example 8 - Manufacture of Vertically Graded Macroporous Scaffold

A macroporous scaffold was made according to Example 1 except that two differently sized populations of paraffin beads were sequentially used to fill the polyethylene tube. Figure 8A-C shows a 2-dimensional Micro CT image of the resultant vertically graded macroporous structure in which the top portion has smaller pores of 300 μm to 400 μm and the bottom portion has larger pores of 600 μm to 700 μm, a top view 3-dimensional Micro CT image of the vertically graded macroporous structure showing the smaller pores at the top surface, and a bottom view 3-dimensional Micro CT image of the structure showing the larger pores at the bottom surface.

Example 9 - Manufacture of Compositionally Graded Macroporous Scaffold

A macroporous scaffold was made according to Example 1 except that two concentrically arranged polyethylene tubes were utilized and different compositions of ceramic material were poured into each tube. The centrally positioned tube contained a ceramic material that was relatively hydroxyapatite (HA) enriched, had a calcium/phosphorus (Ca/P) ratio of about 1.64-1.68:1, and contained titanium oxide. The peripherally positioned tube contained a ceramic material that was relatively tricalcium phosphate (TCP) enriched, had a Ca/P ratio of about 1.48-1.51:1, and did not contain titanium oxide. Figure 9a shows measurements obtained at selected locations in the scaffold. Figure 9b shows the varying composition of the scaffold at each of these selected locations.

As shown in Figure 9b, the Ca/P ratio was higher, between 1.64-1.68:1, in the central HA enriched area of the scaffold compared to between 1.48-1.51:1 in the peripheral areas of the scaffold. Additionally, higher concentrations of titanium, 1.55-1.66:1, were present in the central area and the amount of titanium in the peripheral areas was at or about zero. This result established that there was little movement of slurry components during the template-casting procedure and that the method of the invention may be used to produce compositionally graded ceramic articles.

Example 10 - Controlled Degradation of CaP Scaffolds

Four groups of CaP (β-TCP) scaffolds of the invention were made according to Examples 1 and 7 above to produce (Group A) scaffolds with uniform large pores (between 600 μm and 800 μm) , (Group B) scaffolds with uniform small pores (between 300 μm to 400 μm) , (Group C) radially-graded scaffolds with central small pores and peripheral large pores, and (Group D) radially-graded scaffolds with central large pores and peripheral small pores. Each of the four groups of scaffolds had the same porosity, between 70-73%.

The scaffolds were soaked in Tris buffer (pH 7.4) at 37° C. The dissolution rates of the four groups of scaffolds were measured for a period of 4 weeks. Data is shown in Figure 10. As shown in Figure 10, the graded CaP scaffolds with central big pores and peripheral small pores (Group D) exhibit significantly higher dissolution rate than those with uniform small pores (Group B) and the other graded scaffolds with central small pores and peripheral big pores (Group C) in the course of dissolution. In addition, the scaffolds with uniform big pores have the significantly lowest dissolution rate compared to other groups. No significant difference in dissolution rate was noted between the scaffolds with uniform small pores and the graded scaffolds with central small pores and peripheral big pores. It is postulated that the higher dissolution rate of the scaffolds with uniform small pores is due to their higher surface area compared to those with uniform big pores. It is also postulated that a tension stress caused by the graded architecture resulted in a high dissolution rate for the graded scaffolds of Groups C, those with central big pores and peripheral small pores.

Scaffolds of Group C and Group D were immersed in acidic buffer media (pH 3) . The degradation pattern of these scaffolds is shown in Figure 11. The scaffold regions with faster dissolution were observed to be the regions with smaller pores regardless of the location of the regions.

The in vivo biodegradation of the scaffolds was evaluated. Scaffolds were implanted subcutaneously into mice and the morphology changes were evaluated using non- decalcified histological samples. The results, shown in

Figure 12, were similar to the in vitro study above. Figure 12, panels A-D, show 4 different CaP scaffolds one month after implantation. Panel A is a scaffold from Group A with uniform large pores of 600 to 800 μm. Panel B is a scaffold from Group B with uniform small pores of 350 to 500 μm. Panel C is a graded scaffold from Group C with central small pores of 350 to 500 μm and peripheral large pores of 600 to 800 μm. Panel D is a graded scaffold from Group D with central large pores of 600 to 800 μm and peripheral small pores of 350 to 500 μm. Consistent with dissolution results obtained in vitro,

Figure 12 shows that one month after implantation in vivo, the regions of the scaffolds with smaller pores had also degraded more rapidly than had the regions with larger pores. The results demonstrate that architecture of the scaffolds can be used to guide spatial biodegradation in vivo. Example 11 - Protein Loading of CaP Scaffolds

The effects of varying the loading method and of varying pore size of scaffolds on the elution profile of proteins was evaluated utilizing bovine serum albumin (BSA) and Bone Morphogenetic Protein-2 (BMP-2) .

The BSA was loaded onto the porous scaffolds in two ways, by a pipette method and by an immersion method. In the pipette method, a BSA solution was pipetted directly into the porous scaffolds. In the immersion method, the porous scaffolds were immersed into a BSA protein solution having the same concentration as was used for the pipette method. The subsequent elution profile for the protein was then evaluated. The pipette method resulted in consistent BSA loading and elution profiles for porous scaffolds of all pore sizes. In contrast, the immersion method produced significant differences in loading and elution for porous scaffolds that was dependent on the pore size in the scaffold.

The immersion method was used to load BMP-2 onto the porous scaffolds. Figure 13 shows that the immersion method resulted in a pore size dependent initial loading for BMP-2 that was similar to that for the loading of BSA. Figure 14 shows that the elution profiles over a 21 day period can be regulated by varying scaffold pore size when using the immersion method of loading protein.

Example 12 - CaP Scaffolds Loaded with BMP-2

A study was performed to determine if varying the architecture of CaP scaffolds would have a temporal and/or spatial effect on BMP-2 induced osteogenesis. BMP-2 was loaded into the scaffolds by the immersion method as described above and the scaffolds were implanted subcutaneously into mice. One month after implantation, BMP-2 induced ectopic bone formation was evaluated by micro CT scan and histomorphometry . Figure 15 shows the BMP-2 induced ectopic bone formation in the non-decalcified porous CaP scaffolds at one month after implantation. Micro CT images in panels Al, Bl, Cl, and Dl clearly demonstrate that the porous scaffolds are filled with substances. The histology pictures in panels A2, B2, C2, and D2 confirm that the substance filling the porous scaffolds is newly formed bone. When viewed at higher magnification (not shown) , it was clear that the newly formed bone seamlessly contacts the scaffolds and fills the interconnective pores.

Table 2 lists the histomorpometrical results of bone formation .

Table 2

As shown in Table 2, graded scaffolds with central 600- 800 μm pores and peripheral 350-500 μm pores exhibited significantly greater bone formation compared to uniform scaffolds with 600-800 μm pores (P=O.04089) and graded scaffolds with central 350-500 μm pores and peripheral 600-800 μm pores (P=O.03345) . The uniform scaffolds with 350-500 μm pores did not exhibit significantly different bone formation as compared to uniform scaffolds with 600-800 μm pores (P=O.53853) and to graded scaffolds with central 350-500 μm pores and peripheral 600-800 μm pores (P=O.69125) . These studies indicate that an optimum architecture for CaP scaffolds for induction of osteogenesis may be the graded scaffold with central large pores and peripheral small pores.

Example 13 - Scaffold-aided Bone Healing The porous CaP scaffold of the invention in the presence and absence of recombinant human BMP-2 (rhBMP-2) was evaluated for the ability to enhance bone formation and healing using an accepted rabbit radius critical sized bone defect model. Porous CaP scaffolds loaded with BMP-2 were implanted into a 1.5 cm bone defect in the right radii of New Zealand rabbits, and porous CaP scaffolds without BMP were implanted into a similar defect in the left radii as a control.

It was demonstrated that both the porous scaffolds in the presence and absence of BMP-2 aided bone healing as determined at one month after implantation. Figure 16 shows the radiographic observation of scaffold-aided bone healing at 2 weeks (panel A) and one month (panel B) following implantation. As shown in micro CT images of scaffold-aided bone healing one month after implantation in figure 17, new bone formation is visible among the pores of the scaffolds.

All cited publications, patents, and patent applications are herein incorporated by reference in their entirety.

Further modifications, uses, and applications of the invention described herein will be apparent to those skilled in the art. It is intended that such modifications be encompassed in the above description and in the following claims .

Claims

1. A method for making a ceramic article comprising obtaining a ceramic composition containing a liquid, forming the ceramic composition into a desired shape, hardening the shaped ceramic composition by extracting the liquid of the ceramic composition by exposing the ceramic composition to a solvent in which the liquid is soluble and thereby removing liquid from the ceramic composition, and solidifying the ceramic composition to obtain the ceramic article.

2. The method of claim 1 wherein the liquid contained in the ceramic composition is water.

3. The method of claim 1 wherein the shaped ceramic composition is porous.

4. The method of claim 3 wherein the ceramic composition is shaped by casting onto a replica and wherein the material from which the replica is made is not soluble in the liquid solvent.

5. The method of claim 4 wherein the replica is a negative replica.

6. The method of claim 5 wherein the negative replica is composed of a multiplicity of discrete elements of a sacrificial porogen.

7. The method of claim 6- wherein the elements of the sacrificial porogen are organized into zones that differ based on porogen size.

8. The method of claim 6 wherein the sacrificial porogen is selected from the group consisting of waxes, salts, sodium hydroxide, sugars, gelatins, naphthalene, and polymers.

9. The method of claim 6 wherein the sacrificial porogen is paraffin.

10. The method of claim 9 wherein the liquid solvent is alcohol .

11. The method of claim 10 wherein the alcohol is ethanol.

12. The method of claim 1 wherein the ceramic composition comprises a calcium phosphate ceramic material.

13. The method of claim 12 wherein the calcium phosphate is hydroxyapatite, tricalcium phosphate, or a mixture of hydroxyapatite and tricalcium phosphate.

14. The method of claim 1 wherein the hardening of the ceramic composition is by sequentially exposing the ceramic composition to increasing concentrations of the solvent.

15. The method of claim 4 wherein the replica is removed from the ceramic composition before the ceramic composition is solidified.

16. A ceramic article made by the method of claim 1.

17. The ceramic article of claim 16 which is porous.

18. A ceramic article made by the method of claim 5.

19. A ceramic article made by the method of claim 6.

20. A ceramic article made by the method of claim 14.

21. A porous ceramic article containing pores wherein the pores are 100% interconnected.

22. The porous ceramic article of claim 21 which has a porosity between 60% and 95%.

23. A porous ceramic scaffold that has been manufactured from a liquid-containing ceramic composition by a process that includes a hardening step that comprises exposing the ceramic composition to a solvent in which the liquid of the ceramic composition is soluble at a concentration and for a time sufficient to remove liquid from the ceramic composition and which hardening step is followed by solidifying the ceramic composition to obtain the ceramic scaffold.

24. The porous ceramic scaffold of claim 23 wherein the hardening step is a sequential exposure of the ceramic composition to increasing concentrations of the liquid solvent.

25. The porous ceramic scaffold of claim 23 wherein the porosity of the scaffold is at least 70%.

26. The porous ceramic scaffold of claim 23 wherein the porosity of the scaffold varies between zones of the scaffold.

27. The porous ceramic scaffold of claim 23 wherein the composition of the scaffold varies between zones of the scaffold.

28. The porous ceramic scaffold of claim 23 which is a calcium phosphate ceramic scaffold.

29. The porous ceramic scaffold of claim 23 which comprises solid struts between the pores.

30. A method for repairing a bone defect comprising implanting a solvent-hardened porous ceramic scaffold into a defect of a bone within a patient and maintaining the scaffold at the site of implantation for a time sufficient to permit bony tissue from the bone of the patient to grow in the pores of the scaffold.

31. The method of claim 30 wherein the calcium phosphate comprises tri-calcium phosphate and/or hydroxyapatite .

32. The method of any of claims 30 to 31, wherein the scaffold is loaded with a growth factor.

33. The method of claim 32 wherein the growth factor is bone morphogenic protein.