Bioceramics: from Concept to Clinic

Bioceramics: From Concept to Clinic

Larry L. Hench*

Department of Materials Science and Engineering, University of Florida, Gainesville, Florida 3261 1

Ceramics used for the repair and re- mechanical reliability under load is construction of diseased or damaged still needed. [Key words: bioceramics, parts of the musculo-skeletal sys- structure, dental ceramics, interfaces, tem, termed bioceramics, may be biomechanics.] inert (alumina, zirconia), resorbable (tricalcium phosphate), bioactive (hydroxyapatite, bioactive glasses, and glass-ceramics), or porous for tissue 1. Introduction ingrowth (hydroxyapatite-coated met- MANYmillennia ago, the discovery of als, alumina). Applications include re- human kind that fire would irreversibly placements for hips, knees, teeth, transform clay into ceramic pottery led tendons, and ligaments and repair for eventually to an agrarian society and periodontal disease, maxillofacial re- an enormous improvement in the qual- construction, augmentation and stabi- ity and length of life. Within the last lization of the jaw bone, spinal fusion, four decades another revolution has and bone fillers after tumor surgery. occurred in the use of ceramics to im- Carbon coatings are thromboresistant prove the quality of life. This revolution and are used for prosthetic heart is the innovative use of specially de- valves. The mechanisms of tissue signed ceramics for the repair and re- bonding to bioactive ceramics are be- construction of diseased or damaged ginning to be understood, which can parts of the body. Ceramics used for result in the molecular design of bio- this purpose are termed bioceramics. ceramics for interfacial bonding with Bioceramics can be single crystals hard and soft tissues. Composites are (sapphire), polycrystalline (alumina or being developed with high toughness hydroxyapatite (HA)), glass (Bio- and elastic modulus match with bone. glass@*),glass-ceramics (Ceravital@+ Therapeutic treatment of cancer has or A/W glass-ceramic), or composites been achieved by localized delivery of (stainless-steel-fiber-reinforced Bio- radioactive isotopes via glass beads. glass@ or polyethylene-hydroxyapatite Development of standard test methods (PE-HA)). for prediction of long-term (20-year) Ceramics and glasses have been used for a long time in the health-care R E Newnham-contributing editor industry for eye glasses, diagnostic in- struments, chemical ware, thermome- Manuscript No. 196865 Received March 3, 1991; ters, tissue culture flasks, and fiber approved April 23, 1991. optics for endoscopy. Insoluble porous Supported by U.S. Air Force Office of Scientific Research under Contract No F496290-88-C-0073. glasses have been used as carriers for The conducted research was reviewed and ap enzymes, antibodies. and antigens proved by the University of Florida Institutional since they have several advantages, Animal Care and Use Committee to ensure that notably resistance to microbial attack, the research procedures adhered to the standards set forth in the Guide for the Care and Use of pH changes, solvent conditions, tem- Laboratory Animals (NIH Publication 85-23) as perature, and packing under the high promulgated by the Committee on Care and Use pressure which is required for rapid of Laboratory Animals of the Institute of Labora- tory Animal Resources. Commission on Life Sci- ences, National Research Council. *University of Florida, Gainesville, FL %rnber, American Ceramic Society. 'Leitz Gmbh, Wetrlar. FRG

1487 1488 Journal of the American Ceramic Society - Hench Vol. 74, No. 7

flow.' Ceramics are also widely used in formation of an interfacial bond of im- dentistry as restorative materials, gold plants with bone (Fig. l(b)).4Figure l(b) porcelain crowns, glass-filled ionomer will be discussed in more detail in Sec- cements, dentures, etc. In these appli- tion VIII. cations they are called dental ceram- The relative level of reactivity of an ics as discussed by Preston.' implant influences the thickness of the This review is devoted to the use of interfacial zone or layer between the bioceramics as implants to repair material and tissue. Analysis of failure parts of the body, usually the hard tis- of implant materials during the last sues of the musculo-skeletal system, 20 years generally shows failure origi- such as bones or teeth, although a nating from the biomaterial-tissue in- brief review of the use of carbon coat- terfa~e.~,~When a biomaterial is nearly ings for replacement of heart valves is inert (type 1 in Table I1 and Fig. 1) and Table I. Types of Implant-Tissue also included. Dozens of ceramic com- the interface is not chemically or biologically bonded, there is relative Response positions have been tested;',3 however, few have achieved human clinical ap- movement and progressive develop- If the material is toxic, the surrounding plication. It is now known that clinical ment of a nonadherent fibrous capsule tissue dies. success requires the simultaneous in both soft and hard tissues. Move- If the material is nontoxic and bio- achievement of a stable interface with ment at the biomaterial-tissue inter- logically inactive (nearly inert), a connective tissue and a match of the face eventually leads to deterioration in fibrous tissue of variable thickness mechanical behavior of the implant with function of the implant or the tissue at forms. the tissue to be replaced. the interface or both. The thickness of If the material is nontoxic and bio- the nonadherent capsule varies greatly II. Twes-. of Bioceramics-Tissue depending upon both material (Fig. 2) logically active (bioactive), an inter- Attachment facial bond forms. and extent of relative motion. The mechanism of tissue attachment The fibrous tissue at the interface with dense, medical-grade alumina im- the surrounding tissue replaces it. response at the implant interface.' No plants can be very Consequently, material implanted in living tissues is as discussed later, if alumina implants inert; all materials elicit a response are implanted with a very tight me- from living tissues. The four types of re- chanical fit and are loaded primarily in sponse (Table I) allow different means compression, they are successful clini- of achieving attachment of prostheses cally. In contrast, if a type 1, nearly to the musculo-skeletal system. Table II inert, implant is loaded such that inter- summarizes the attachment mecha- facial movement can occur, the fibrous nisms, with examples. capsule can become several hundred A comparison of the relative chemi- micrometers thick and the implant cal activity of these different types of loosens quickly. Loosening invariably bioceramics is given in Fig. 1. The leads to clinical failure, for a variety of relative reactivity shown in Fig. I(a) reasons, including fracture of the im- correlates very closely with the rate of plant or the bone adjacent to the implant. Bone at an interface with Table II. Types of Bioceramics a type 1, nearly inert, implant is very -Tissue Attachment and Bioceramic Classification often structurally weak because of Type of disease, localized death of bone (es- bioceramic Type of attachment Example pecially if so-called bone cement, 1 Dense, nonporous, nearly inert Al2O3 (single crystal and poly(methy1 methacrylate (PMMA) is ceramics attach by bone polycrystalline) used), or stress shielding when the growth into surface higher elastic modulus of the implant irregularities by cementing prevents the bone from being loaded the device into the tissues, properly. or by press fitting into a The concept behind nearly inert, defect (termed morphologi- microporous bioceramics (type 2 in cal fixation). Table II and Fig. 1) is the ingrowth of 2 For porous inert implants A1203 (porous polycrystalline) tissue into pores on the surface or bone ingrowth occurs, Hydroxyapatite-coated porous throughout the implant, as originated by which mechanically metals Hulbert et aL3 many years ago. The in- attaches the bone to the creased interfacial area between the material (termed biological implant and the tissues results in an fixation). increased inertial resistance to move- 3 Dense, nonporous, surface- Bioactive glasses ment of the device in the tissue. The reactive ceramics, glasses, Bioactive glass-ceramics interface is established by the living and glass-ceramics attach Hydroxyapatite tissue in the pores. Figure 3 shows liv- directly by chemical ing bone grown into the pores of an bonding with the bone alumina bioceramic. This method of (termed bioactive fixation). attachment is often termed biological 4 Dense, nonporous (or porous), Calcium sulfate (plaster of Paris) fixation. It is capable of withstanding resorbable ceramics are Tricalcium phosphate more complex stress states than type 1 designed to be slowly Calcium phosphate salts implants, which achieve only morpho- replaced by bone. logical fixation." The limitation associated with type 2 porous implants, however, is that, for the tissue to remain July 1991 Bioceramics: From Concept to Clinic 1489 viable and healthy, it is necessary for which have evolved over millions of the pores to be greater than 100 to years. Complications in the develop- 150 pm in diameter (Fig. 2). The large ment of resorbable bioceramics are interfacial area required for the porosity (1) maintenance of strength and the is due to the need to provide a blood stability of the interface during the supply to the ingrown connective tissue. degradation period and replace- Vascular tissue does not appear in ment by the natural host tissue and pores which measure less than 100 pm. (2) matching resorption rates to the If micromovement occurs at the inter- repair rates of body tissues (Fig. I(a))> face of a porous implant, tissue is which themselves vary enor mousl y. damaged, the blood supply may be cut Some dissolve too rapidly and some off, tissues die, inflammation ensues, too slowly. Because large quantities of and the interfacial stability can be de- material may be replaced, it is also es- stroyed. When the material is a metal, sential that a resorbable biomaterial the large increase in surface area can consists only of metabolically accept- provide a focus for corrosion of the im- able substances. This criterion implant and loss of metal ions into the poses considerable limitations on the tissues, which may cause a variety of compositional design of resorbable medic a I pro b I em s These poten t ia I biomaterials. Successful examples problems can be diminished by using are resorbable polymers such as a bioactive ceramic material such as HA as a coating on the porous metal, as first shown by Ducheyne et a/.’ The HA coating also speeds the rate of bone formation in the pores. However, the fraction of large porosity required for bone growth in any material de- grades the strength of the material. inert) Consequently, this approach to solving interfacial stability is best when used as porous coatings or when used as unloaded space fillers in tissues. Resorbable biomaterials (type 4 in Table II and Fig. 1) are designed to degrade gradually over a period of time and be replaced by the natural host tis~ue.’~-’~This leads to a very thin or nonexistent interfacial thickness (Fig. 2). This is the optimal solution to the problem of biomaterials if the re- 1 r I I I III I I I 11I I I T I I I 111 quirements of strength and short-term 3 10 100 1000 performance can be met. Natural tis- Implantation time (d) sues can repair themselves and are Fig. 1. Bioactivity spectrum for various biocerarnic implants gradually replaced throughout life by a (a) relative rate of bioreactivity and (b) time dependence of forma- continual turnover of cell populations. tion of bone bonding at an implant interface ((A) 45S5 Bioglass@ As we grow older, the replacement of (6) KGS CeravitaP (C) 5584 3 Bioglass@ (D) A/W glass-ceramic cells and tissues is slower and less ef- (E) HA (F) KGX Ceravitala’, and (G) AI2O3-Si3N4) ficient, which is why parts “wear out,” unfortunately some faster than others. Thus, resorbable biomaterials are based on the same principles of repair

Interfacial thickness (pm) Initial Material interface Bone 1000 100 10 I 0 1 10 100 1000

I 1000 100 10 1 0 1 10 100 1000 Fig. 2. Comparison of interfacial thickness of reaction layer of Fig. 3. lngrowth of bone within >100-pm pores of alumina bioce- bioactive implants or fibrous tissue of inactive bioceramics in bone ramic (Photograph courtesy of S Hulbert ) 1490 Journal of the American Ceramic Society - Hench Vvl. 74, No. 7 poly(lactic acid)-poly(glyco1ic acid) 111. Nearly Inert Crystalline (PLA-PGA) used for sutures, which are Bioceramics metabolized to carbon dioxide and High-de nsi t y, hi gh-pu r it y ( > 9 9.5%) water and therefore are able to func- alumina (a-Al2O3)was the first bioce- tion for a period and then dissolve and ramic widely used clinically. It is used disappear. Porous or particulate Cal- in load-bearing hip prostheses and cium phosphate ceramic materials dental implants because of its combi- such as tricalcium phosphate (TCP) nation of excellent corrosion resistance, are successful materials for resorbable good biocompatibility, high wear re- hard-tissue replacements when only sistance, and high ~trength.’.~~’~-‘~ low mechanical strength is required, Although some dental implants are such as in some repairs of the jaw or single-crystal ~apphire,’~most alumina head. devices are very-fine-grained polycrys- Another approach to the solution of talline a-Al,O,. A very small amount of the problems of interfacial attachment magnesia (<0.5%) is used as an aid to is the use of bioactive materials (type 3 sintering and to limit grain growth dur- in Table II and Fig. 1). The concept of ing sintering. bioactive materials is intermediate be- Strength, fatigue resistance, and tween resorbable and bi~inert.’,~’~A fracture toughness of polycrystalline bioactive material is one that elicits a a-Al2O3are a function of grain size and specific biological response at the inter- percentage of sintering aid, i.e., purity. face of the material which results in the Alumina with an average grain size of formation of a bond between the tissues <4 prn and >99.7% purity exhibits and the material (shown first in 1969).’4 good flexural strength and excellent This concept has now been expanded compressive strength. These and other to include a large number of bioactive physical properties are summarized in materials with a wide range of rates of Table Ill6 with the International Stand- bonding and thickness of interfacial ards Organization (ISO) requirements bonding layers (Figs. 1 and 2). They in- for alumina implants. Extensive testing ~Iude~.~bioactive glasses such as Bio- has shown that alumina implants which glass? bioactive glass-ceramics such meet or exceed IS0 standards have as Ceravitalo, A/W glass-ceramic, or excellent resistance to dynamic and machineable glass-ceramics; dense HA impact fatigue and also resist subcriti- such as durapatite or Calcitite@*, or cal crack growth.20An increase in the bioactive composites such as Palavi- average grain size to >7 pm can de- tat@§,stainless-steel-fiber-reinforced crease mechanical properties by about BioglassB; and PE-HA mixtures. All of 20%. High concentration of sintering these bioactive materials form an inter- aids must be avoided because they re- facial bond with adjacent tissue. How- main in the grain boundaries and de- ever, the time dependence of bonding, grade fatigue resistance, especially in the strength of the bond, the mecha- a corrosive physiological environment.’ nism of bonding, and the thickness of Methods exist for lifetime predictions the bonding zone differ for the various and statistical design of proof tests for materials. load-bearing ceramics. Applications of It is important to recognize that rela- these techniques show that specific tively small changes in the composition prosthesis load limits can be set for an of a biomaterial can affect dramatically alumina device based upon the flexural whether it is bioinert, resorbable, or strength of the material and its use en- bioactive. These compositional effects vironment.*’ Load-bearing lifetimes of on surface reactions are discussed in 30 years at 12000-N loads, similar to Section V. those expected in hip joints, have been predicted.6 Results from aging and fatigue studies show that it is essential that alumina implants be produced at *Calatek, lnc , San Diego, CA ‘Leitz Gmbh the highest possible standards of qual-

Table 111. Physical Characteristics of Alumina and Partially Stabilized Zirconia (PSZ) Bioceramics IS0 alumina Cortical Cancellpus High alumina ceramics* standard 6474 PSZ* bone+ bone Content (percent by weight) A1203 > 998 A1203 2 9950 Zr02 > 97 Density (g/cm3) > 393 z 390 5 6-6 12 1 6-2 1 Average grain size (pm) 3-6 c7 1 Surface roughness, R, (pm) 0 02 0 008 Hardness (Vickers), HV 2300 > 2000 1300 Compressive strength (MPa) 4500 2-12 Bendina strenath IMPa) 550 400 1200 50-150 (aftertesting in Ringer’s solution) Young’s Modulus (GPa) 380 200 7-25 0.05-0.5 Fracture toughness, Kc (MPa. m-1’2) 5-6 15 2-12 Slow crack growth, n (unitless) 30-52 65 *Reference 6 ‘References 120 and 121 *Reference 122 July 1991 Bioceramics: From Concept to Clinic 1491 ity assurance, especially if they are to mina. Christel et a/.6 report that stress be used as orthopedic prostheses in shielding may be responsible for can- younger patients (-40years old). cellous bone atrophy and loosening of Alumina has been used in orthope- the acetabular cup in older patients dic surgery for nearly 20 years, moti- with senile osteoporosis or rheumatoid vated largely by (1) its excellent type 1 arthritis. Consequently, it is essential biocompatibility and very thin capsule that the age of the patient, nature of the formation (Fig. 2) which permits ce- disease of the joint, and biomechanics rnentless fixation of prostheses3 and of the repair be considered carefully be- (2) its exceptionally low coefficients of fore any prosthesis is used, including friction and wear rate^.^,^ those made from alumina ceramics. In The superb tribiologic properties the United States, the primary use of (friction and wear) of alumina occur only alumina is for the ball of the hip joint when the grains are very small (14pm) (Fig. 5), with the acetabular component and have a very narrow size distribu- being ultrahigh-molecular-weight PE. tion. These conditions lead to very low Other clinical applications of alumina surface roughness values (R,50.02 pm, prostheses, reviewed by Hulbert et a/.,3 Table Ill). If large grains are present, include knee prostheses, bone screws, they can pull out and lead to very rapid alveolar ridge (jaw bone) and maxillo- wear because of local dry friction and facial reconstruction, ossicular (middle abrasion caused by the alumina grains ear) bone substitutes, keratoprosthe- in the joint-bearing surfaces.6 ses (corneal replacements), segmental Alumina on alumina load-bearing bone replacements, and blade and wearing surfaces, such as in hip pros- screw and post-type dental implants. theses, must have a very high degree of sphericity produced by grinding and IV. Porous Ceramics polishing the two mating surfaces to- The potential advantage offered by a gether. The alumina ball and socket in porous ceramic implant (type 2, a hip prosthesis are polished together Table II, Figs. 1 and 2) is its inertness and used as a pair. The long-term coef- combined with the mechanical stability ficient of friction of an alumina-alumina of the highly convoluted interface devel- joint decreases with time and ap- oped when bone grows into the pores proaches the values of a normal joint. of the ceramic. Mechanical require- This leads to wear of alumina on alu- ments of prostheses, however, severely mina articulating surfaces that are restrict the use of low-strength porous nearly 10 times lower than metal-PE ceramics to low-load- or non-load- surfaces (Fig, 4). bearing applications. Studies show Low wear rates have led to wide- that, when load bearing is not a pri- spread use in Europe of alumina non- mary requirement, nearly inert porous cemented cups press fitted into the ceramics can provide a functional im- acetabulum (socket) of the hip. The plant,1.3.22-24 When pore sizes exceed cups are stabilized by bone growth into 100 pm, bone will grow within the inter- grooves or around pegs. The mating connecting pore channels near the sur- femoral ball surface is also of alumina face and maintain its vascularity and which is bonded to a metallic stem. long-term viability (Fig. 3). In this man- Long-term results in general are excel- ner the implant serves as a structural lent, especially for younger patients. bridge and model or scaffold for bone However, Christel et a/.6 caution that formation. The microstructures of cer- stress shielding of the bone can occur. tain corals make an almost ideal in- This is due to the high Young's modu- vestment material for the casting of lus of alumina (Table Ill), which prevents the bone from being loaded, a requirement for bone to remain healthy and strong. The Young's modulus of cortical bone ranges between 7 and 25 GPa (as discussed in Section X) which is 10 to 50 times lower than alu-

g 0'15 Friction I .- [ - r:

lb 100 id00 1 &oo Testing time (h)

Fig. 4. Time dependence of (-) coefficient of friction and (---) Fig. 5. Medical-grade alumina used as femoral balls in total hip index of wear of alumina-alumina versus metal-PE hip joint (in vitro replacement Note three alternative types of metallic stems used for testing) morphological fixation (Photograph courtesy of J Parr) 1492 Journal of the American Ceramic Society - Hench Vol. 74, No. 7

structures with highly controlled pore on decreasing the strength become sizes White et a/’3 developed the re- much more important than for dense, plamineform process to duplicate the nonporous materials The in vitro aging porous microstructure of corals that of porous alumina in saline solution for have a high degree of uniform pore periods of 4 weeks as well as in VIVO size and interconnection The first step testing for up to 3 months reduces is to machine the coral with proper strength of porous alumina samples microstructure into the desired shape by 35% to 40% Permeation into the The most promising coral genus, Por- micropores of seemingly dense alumina ites has pores with a size range of 140 can also result in marked reductions in to 160 pm, with all the pores intercon- tensile strength Another porous, high- nected ‘’ Another interesting coral strength ceramic calcia-stabilized genus, Goniopora, has a larger pore zirconia, also undergoes significant size, ranging from 200 to 1000 pm decreases in strength with time as The machined coral shape is fired to do calcium aluminates ’ Aging of porous drive off carbon dioxide from the lime- ceramics with their subsequent de- stone (CaCO,) forming calcia, while crease in strength poses questions as maintaining the microstructure of the to the successful long-term application original coral The calcia structure of porous materials unless they are de- serves as an investment material for signed to be resorbable, eg , are made forming the porous material After the of calcium salts such as TCP desired material is cast into the pores, the calcia is easily removed from the material by dissolving it in dilute hy- V. Bioactive Glasses and drochloric acid The primary advan- Glass-Ceramics tages of the replamineform process are Certain compositions of glasses, ce- that the pore size and microstructure ramics, glass-ceramics, and com- are uniform and controlled, and there posites have been shown to bond to is complete interconnection of the bone,’ 3.4 5.79-14 25 30-40 Th ese materials pores Replamineform porous materials have become known as bioactive ce- of a-A1203,titania, calcium phosphates, ,4.5.25.41 Some even more special- polyurethane silicone rubber, PMMA, ized compositions of bioactive glasses and Co-Cr alloys have been used as will bond to soft tissues as well as bone implants with the calcium phos- bone.354’A common characteristic of phates being the most acceptable 22 25 bioactive glasses and bioactive ceram- Porous ceramic surfaces can also ics is a time-dependent, kinetic modifi- be prepared by mixing soluble metal or cation of the surface that occurs upon salt particles into the surface The implantati~n.~.~The surface forms a bio- pore size and structure are determined logically active hydroxycarbonate by the size and shape of the soluble apatite (HCA) layer which provides the particles that are subsequently re- bonding interface with tissues. The moved with a suitable etchant The HCA phase that forms on bioactive porous surface layer produced by this implants is equivalent chemically and technique is an integral part of the un- structurally to the mineral phase in derlying dense ceramic phase bone. It is that equivalence which is re- Materials such as alumina may also sponsible for interfacial bonding. be made porous by using a suitable Materials that are bioactive develop foaming agent that evolves gases dur- an adherent interface with tissues that ing heating Porous alumina and cal- resists substantial mechanical forces. cium aluminates used by Hulbert and In many cases the interfacial strength ~~lleagues~~~~~in some of their bone in- of adhesion is equvalent to or greater growth studies were produced by mix- than the cohesive strength of the im- ing powdered calcium carbonate with plant material or the tissue bonded to fine alumina powder A firing time of the bioactive implant. Figures 6 and 7 20 h at approximately 1450” to 1500°C show bioactive implants bonded to produced a foamed material with pore bone with adherence at the interface size and volume fraction (33% to 48% sufficient to resist mechanical fracture. of porosity) determined by the size and Failure occurs either in the implant concentration of the original calcium (Fig. 6) or in the bone (Fig. 7) but al- carbonate particles most never at the interface. Porous materials are weaker than Bonding to bone was first demon- the equivalent bulk form As the poros- strated for a certain compositional ity increases the strength of the ma- range of bioactive glasses which con- terial decreases rapidly as indicated tained SO,, Na20, CaO, and P20, in by the Ryskewitch equation specific proportions (Table IV).I6 There were three key compositional features = aOe-cp (1) u to these glasses that distinguished where ~r is strength v0 IS strength at them from traditional NanO-CaO-SiOn zero porosity, c is constant, and p is glasses: (1) less than 60 mol% SO2, porosity (2) high-NanO and high-CaO content, Much surface area is also exposed, and (3) high-CaO/P,O, ratio. These so that the effects of the environment compositional features made the sur- July 1991 Bioceramics: From Concept to Clinic 1493

face highly reactive when exposed to region B (such as window, bottle, or an aqueous medium. microscope slide glasses) behave as Many bioactive silica glasses are type 1, nearly inert materials and elicit based upon the formula called 45S5 a fibrous capsule at the implant-tissue (signifying 45 wt% SO,,S as the net- interface. Glasses within region C are work former, and a 5 to 1 molar ratio resorbable and disappear within 10 to of Ca to P). Glasses with substantially 30 d of implantation. Glasses within lower molar ratios of Ca to P (in the region D are not technically practical form of CaO and P205)do not bond to and therefore have not been tested as bone.43 However, substitutions in the implants. 45% formula of 5 to 15 wt% BZO3for The collagenous constituent of soft SiO, or 12.5 wt% CaF, for CaO or tissues can strongly adhere to the crystallizing the various bioactive glass bioactive silica glasses which lie within compositions to form glass-ceramics the compositional range (dashed line) has no measurable effect on the ability shown in Fig. 8. Figure 9 shows colla- of the material to form a bone bond.43 gen from a 10-d in vitro test-tube ex- However, addition of as little as 3 wt% periment bonded to a 45% Bioglasso A1203 to the 45S5 formula prevents surface by agglomerates of HCA crys- bonding.53'43-46 tallites growing on the surface. The The compositional dependence (in collagen fibrils are woven into the weight percent) of bone bonding and interface by growth of the HCA layer soft-tissue bonding for the Na,O- (Fig. 9(a)).The dense HCA-collagen CaO-P,O,-SiO, glasses is illustrated agglomerates (Fig. 9(b)) mimic the in Fig. 8. All glasses in Fig. 8 contain a nature of bonding between tendons constant 6 wt% of P,O,. Compositions and ligaments, composed entirely of in the middle of the diagram (region A) collagen fibrils, and bone which is form a bond with bone. Consequently, a composite of HCA crystals and region A is termed the bioactive-bone- collagen.4748The composite interface bonding boundary. Silica glasses within composed of HCA collagen on the

Fig. 6. Fracture of (BGC) 4585 Bioglass@-ceramic segmental Fig. 7. Interfacial adherence of A/W bioactive glass-ceramic is bone replacement in monkey due to impact torsional loading 76 Note stronger than either (B) bone or (AIW) implant (Photograph cour (I) bonded interface (Photograph courtesy G Piotrowski ) tesy T Yamamura)

Table IV. Composition of Bioactive Glasses and Glass-Ceramics (wt%) 4555 45S54F 45B15S5 52S46 55S43 KGC KGS KGy213 AIW MB Component Bioglass" Bioglass@ Biogiassa Bioglassm Boglass@ Ceravilal@ CeravttaP Ceravilaia glass-ceramlc glass-ceramic s45P7 90, 45 45 30 52 55 462 46 38 34 2 19-52 45 pzo5 6 6 6 6 6 16 3 4-24 7 CaO 24 5 147 245 21 195 202 33 31 44 9 9-3 22 Ca(P03), 25 5 16 13 5 CaF, 98 05 MgO 29 46 5-15 MgF2 Na,O 24 5 245 245 21 195 48 5 4 3-5 24 Kz0 04 3-5 A1203 7 12-33 B~03 15 2 Ta2O5/TiO2 65 Structure Glass and Glass Glass Glass Glass- Glass- Glass- Glass Glass- glass- ceramic ceramic ceramic ceramic ceramic c e r a mi c Reference 14 14, 56 57 58 44 44 5 5 5 36 32 54 1494 Journal of the American Ceramic Society - Hench Vol. 14, No. 7

I-, bioactive glass is approximately 30 to posed of apatite (Ca,,(PO4),(OH,F2)) 60 pm of the 100- to 200-pm total in- and wollastonite (CaO . Si02) crystals terfacial thickness (Fig 2) This junc- and a residual Si02 glassy matrix, tion thickness is equivalent to that at termed AIW glass-ceramic by the naturally occurring interfaces where a Kyoto University team in Japan,34.36-40 transition occurs between materials also bonds with bone and has very (tendons and ligaments) with a low high interfacial bond ~trength.~"'Addi- Young's modulus and bone with a tion of A1203or Ti02 to the A/W glass- moderately high Young's modulus The ceramic also inhibits bone bond- thickness of the hard-tissue-bioactive ing, whereas incorporation of a sec- ceramic interfaces is indicated in ond phosphate phase, P-whitlockite Fig 2 for several compositions The (3Ca0. P205),does not. interfacial thickness decreases as the Another rnultiphase bioactive silica- bone-bonding boundary shown in Fig 8 phosphate glass-ceramic containing is approached phlogopite ((Na,K)Mg3(AISi3Ol0)F2),a Gross and co-workers5 31 4y 52 have mica, and apatite crystals, bonds to shown that a range of low-alkali (0 to bone even though A1203 is present in 5 wt%), bioactive silica glass-ceramics the comp~sition.~~However, the A13+ (CeravitaP) also bond to bone They ions are incorporated within the crystal find that small additions of Al2O3, phase and do not alter the surface re- Ta205,Ti02, Sb2O3, or Zr02 inhibit bone action kinetics of the material. An bonding (Table IV, Fig 1) A two-phase advantage of these mica-containing silica-phosphate glass-ceramic com- glass-ceramics, developed by the Freidrich Schiller University, Jena, Fed- eral Republic of Germany, is their easy ma~hinability.~~Additional compositions of bioactive glasses have been developed at Abo Akademi, Turku, Finland, for coating onto dental alloy^.^^,^^ 55 Compositions of these various bioactive glasses and glass-ceramics are compared in Table IV.

VI. Interfacial Reaction Kinetics The literature base for stage 1 (ion exchange) and stage 2 (silica network dissolution) reactions (see Table k) is quite extensive, as reviewed in Refs. 60 to 64. Measurements of stage 3, silica repolymerization, are less exten~ive~~,~~ but fairly conclusive. Surface composi- Fig. 8. Compositional dependence (in weight percent) of bone tional profiles resulting from stage 1 bonding and soft-tissue bonding of bioactive glasses and glass- 4 ceramics All compositions in region A have a constant 6 wt% of through stage reactions have been P205 A/W glass-ceramic has higher P205content (see Table IV for measured for numerous glass com- details) Region E (soft-tissue bonding) is inside the dashed line position~.~~-~~The effects of addition where /+8 ((*) 45% Bioglasse (D) CeravitaP (0) 55843 Bio of sparingly soluble cations, such as glass@,and (---) soft-tissue bonding, /B=100/to5bb) Mg, Ca, and Al, to glasses and/or

(A) (B) Fig. 9. (a) SEM micrograph of collagen fibrils incorporated within the HCA layer growing on a 4585 Bioglass@substrate in vitro (b) Close-up (-11300~)of the HCA crystals bonding to a collagen fibril (Photographs courtesy of C Pantano) July 1991 Bioceramics: From Concept to Clinic 1495 reaction solutions are also well docu- the chemical species in the surface. merited,60.61.63.64 The effects of glass- Figure 10 shows the 45S5 glass surface crystal interfaces on overall reaction before reaction (0 h), after 1 h. and after rates have been measured for simple 2 h. The peak identifications are based systems,5gbut not on complex, multi- upon previous assignments of IR spec- phase bioactive glass-ceramics or tra.7375 The alkali-ion-hydrogen-ion bioactive composites. exchange and network dissolution Table V summarizes the time- (stages 1 and 2 in Table V) very rapidly dependent changes of only a single- reduces the intensity of the Si-0-Na phase amorphous or glassy material. If and Si-0-Ca vibrational modes and one desires to understand a multiphase replaces them with Si-OH bonds with bioactive implant, such as CeravitaP’, one nonbridging oxygen (NBO) ion. A/W glass-ceramic, polycrystalline Alkali content is depleted to a depth sintered HA, or a bioactive composite, >0.5 pm within a few minutes. it is necessary to establish the time- As the stage 1 and 2 reactions con- dependent surface changes for each tinue, the Si-OH single, NBO modes phase and each interface between the are replaced with phases. The biological environment may degrade a phase or an interface H preferentially leading to grain-boundary 0 degradation for certain bioactive glass- I ceramics, as discussed by Gross et aL5 -0-Si-OH I VII. Example of Glass-Surface 0 Reaction Kinetics (4585 Bioglassa) The original bioactive glass 4595 has been used as a baseline for most i.e., Si-2NB0 stretching vibrations surface studies, largely because it is which are in the range of 930 cm-’, de- single phase and has only four compo- creasing to 880 cm-’. By 20 min, the nents (Na,O, CaO, PZO5, Si02). This Si-2NB0 vibrations (Fig. 11) are largely simple composition (Table IV) also has replaced by a new mode assigned to the highest in vivo bioactivity index (I,) the Si-0-Si bond vibration between as discussed later. Recent investiga- two adjacent SiO, tetrahedra. Thus, t ions66.70,71 show quite clearly the reac- this new vibrational mode corresponds tion sequence depicted by Eq. (1) for to the formation of the silica-gel layer 45S5 Bioglassm exposed to tris(hy- by the stage 3 (Table V) polyconden- droxymethy1)aminomethane buffer solu- sation reaction between neighboring tion (TBS) that does not contain Ca or surface silanol groups. This mode de- P ions (TBS). Figures 10 and 11 sum- creases in frequency until it is hidden marize these findings, determined us- by the growing apatite layer after 1 h. ing Fourier transform infrared (FTIR) As early as 10 min, a P-0 bending spectroscopy of the reacted surface vibration associated with formation of as a function of exposure time. All five an amorphous calcium phosphate reaction stages are clearly delineated layer appears. This is due to precipita- by changes in the vibrational modes of tion from solution (stage 4 in Table V).

Table V. Reaction Stages of a Bioactive Implant Stage Reaction 1 Rapid exchange of Na+or KCwith Hf or H30+from solution Si-0-Na++H ‘+OH -+Si-OH’+Na+(solution)+OH-

This stage is usually controlled by diffusion and exhibits a t-’12 dependence 6o 63 2 Loss of soluble silica in the form of SI(OH)~to the solution, resulting from breaking of Si-0-Si bonds and formation of Si-OH (silanols) at the glass solution interface SI-0-SI +H ,O+SI-OH+OH-SI This stage IS usually controlled by interfacial reaction and exhibits a t” dependence606’ 3 Condensation and repolymerization of a SiOp-nch layer on the surface depleted in alkalis and alkaline-earth cations 0 0 00 I I I i O-SI-OH+HO-SI-O+O-S~-OSI-O+H,O I I I I 0 0 00 4 Migration of Caz+and PO:- groups to the surface through the SiOz-rich layer forming a CaO-P205-rich film on top of the Si02-rich layer, followed by growth of the amorphous6768,, Ca0-PZO5-rich film by incorporation of soluble calcium and phosphates from solution 5 Crystallization of the amorphous CaO-P,O, film by incorporation of OH-, CO,’-, or F- anions from solution to form a mixed hydroxyl, carbonate, fluorapatite layer 1496 Journal of the American Ceramic Society - Hench Vol. 74, No. 7

Clark eta/ 67 showed, using Auger elec- ing vibration is strong and exhibits a tron spectroscopy (AES) and Ar-ion- continually decreasing frequency as the beam milling, that, even by 2 min, calcium phosphate-rich layer builds. At calcium and phosphate enrichment oc- about 1.5&0.2h, the P-0 bending vi- curred on the sample surface to a bration associated with the amorphous depth of approximately 20 nm Ogino calcium phosphate layer is replaced eta/ 67 showed that, by 1 h, the calcium with two P-0 modes assigned to phosphate layer grew to 200 nm in crystalline apatite. Concurrent with the thickness The bilayer films of calcium onset of apatite crystallization (stage 5 phosphate on top of polymerized silica in Table V) is the appearance of a C-0 gel, obtained by AES and Ar-ion mill- vibrational mode associated with the ing, are shown in Fig 12 The calcium incorporation of COZ- in the apatite phosphate-rich layer extends to a crystal lattice as described by Kim depth of nearly 08 pm after only 1 h in et a/.7' and Le Geros et a/.75The C-0 rat bone The silica-rich film is already mode decreases in wavenumber as the several micrometers thick Organic HCA layer grows. By 10 h the HCA layer constituents containing C and N are has grown to 4 pm in thickness,67 incoworated in the growing calcium which is sufficient to dominate the FTlR phosphate-rich film. spectra and mask most of the vibra- Within 40 min (Fig 11) the P-0 bond- tional modes of the silica-gel layer or the bulk-glass substrate. By 100 h P-0 stretch 26.55 the polycrystalline HCA layer is thick enough to yield X-ray diffraction (XRD) results showing the primary 26" and 33" 20 peaks with considerable line 21.06 br~adening.'~'~~By 2 weeks the crys- g- talline HCA layer is equivalent to bio- al logical apatites grown in v~vo.~~ Kok~bo~'~~showed that a calcium 5 15.56 phosphate-rich layer is also present at -a, - the bonding interface between the d polycrystalline apatite and wollastonite- 10.07 based A/W glass-ceramic and bone. However, the 30,-rich layer was not present, even though a substantial concentration of soluble Si was lost 4.57 Si-0 stretch to solution. Likewise, Kokubo and 20.29 co-workers6' showed that the calcium

Surface reaction stages Si-0 bend 15.51 2 6 -0 4- al0 5 10.72 960 690 m P-0 - (amorphous) d 1 Si-0-Si stretch 920 650 5.93

880 61 0 1.15 - 3-0 stretch 45.30 -.- 2 5 - 840 :\: T ,570 5 n& IL5 P-0 bend 0- 34.25 - 5 f (crystalline) c - 800 ,530 Z $. 7 m 3 - 5 23.20. Si-0-Si stretch - 760 Si-0-Si bend (tetragonal) ,490 0) c- d 12.15- 720 ,450

1.09- ~ 680 -410 1-1 I I I 11111111111 3 1200 1000 800 600 400 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Wavenum ber log t (s) Fig. 10. FTlR spectra of a 45S5 Bloglass@implant after 0, 1 and Fig. 11. Time-dependent changes in IR vibrations of the surface 2 h in TBS at 37°C (Diagram courtesy of G LaTorre) of 4585 Bioglass@ implant in a 37°C TBS (Diagram courtesy of G LaTorre) July 1991 Bioceramics: From Concept to Clinic 1497 phosphate-rich layer was present for Ceravi t al@-t ype glass-ceram ic s. Hohland et reported a calcium phosphate-rich layer at the bone interface with a glass-ceramic composed of phologopite and apatite crystals. Kokubo and co-~orkers~~~~also demonstrated that CaO-SiOz-based glasses, without phosphate, formed an apatite layer on their surface when exposed for 2 to 30 d in simulated body fluid that contained only 1.OmM HPOi-. The CaO-Si02 glasses were confirmed to bond to living bone by the surface apatite layer.77Previously Ogino, Ohuchi, and the showed that P,O,-free Na,O-SiO, glasses formed an apatite layer on their surface when exposed to an aqueous solution containing calcium and phosphate ions. Recently, Li, Clark, and the author7’ demonstrated that glasses containing primarily SiO,, with only 10 mol% of CaO and Pz05 and no Na20 formed apatite layers in a tris buffer solution. Earlier, Walker” Fig. 12. Eilayer films formed on 4555 Bioglass@after 1 h in rat demonstrated that even nearly pure bone, in vivo (1 h=lO-’ nm) (from Ref. 68) SiOz eventually formed a bone bond if the surface had a very high surface tered, as shown in Fig 1 Thus, the area, >400 m2/g. Unfortunately, the relative bioreacttvity of the material, interface was not analyzed for the also shown in Fig 1, is composition presence of an interfacial apatite layer dependent The level of bioactivity of a which could have been nucleated on specific material can be related4 to the the surface by hydroxylation andlor time for more than 50% of the inter- dissolution of soluble 30,. For years it face to be bonded (tOsbb,where /B= has been known that synthetic HA im- 100/to5bb) Gross, Strunz, and col- plants, which contain no SiO, or alkali leaguess4’ 52 have shown that the initial ions, will bond to bone by forming concentration of cells present at the in- a new epitaxial apatite phase at the terface (stem cells, osteoblasts, chon- interface,’0~8’,82as discussed in a later droblasts, and fibroblasts) varies as a sect ion. function of the fit of the implant and Consequently, we conclude that the condition of the bony defect Con- bioactivity occurs only within certain sequently, all bioactive implants require compositional limits and very specific an incubation period before bone pro- ratios of oxides in the Na20-KzO- liferates and bonds, which is evident in Ca0-Mg0-P,O,-SiO2 systems; how- Fig 1 The length of the incubation ever, the extent of these compositional period at which this process occurs limits and the physical, chemical, and varies over a wide range depending on biochemical reasons for the limits are implant composition, which controls poorly known at present. the kinetics of the surface reactions We do know that, for a bond with tis- (stages 1 to 5 in Table V) For the sues to occur, a layer of biologically ac- material to be bioactive and form an tive HCA must form. This is perhaps the interfacial bond, the time of stages 4 only common characteristic of all the and 5 must match the time of biomin- known bioactive implant materials. It is eralization that normally occurs in vivo the rate of HCA formation (stage 4) If the surface reactions are too rapid, and the time for onset of crystallization the implant is resorbable (type 4, (stage 5) that varies so greatly. When Table II) If the reactions are too slow, the rate becomes excessively slow, no the implant is not bioactive, i e ~ it has bond forms and the material is no a type 1 response longer bioactive. Compositional effects The compositional dependence of lB on stages 4 and 5 appear to be critical (Fig 1) shows that there are isole con- in controlling bonding, nonbonding, or tours within the bioactivity boundary, resorption. Our next step in understand- as shown in Fig 84The change of IB ing is to relate the implant surface re- with Si02/(Na,0+CaO) ratio is very actions to in vivo response. large as the bioactivity boundary is approached, at 60%-StO2lB-0 Addltion VIII. Relation of Surface Kinetics of multivalent ions to a bioactive glass to Rate of Bone Bonding or glass-ceramic will serve to shrink By changing the compositionally the isoIB contours Thus, the isole con- controlled reaction kinetics (Table V), tours decrease as the percentage of the rates of formation of hard tissue at A1203 TanO,, and ZrO, increases in a bioactive implant interface can be al- the material Consequently, the isolB 1498 Journal of the American Ceramic Society - Hench Vol. 74, No. 7

~~~~~~~-~~-m boundary shown in Fig 8 shows and soft-tissue bonding depending on the contamination limit for bioactive the progenitor stem cells in contact glasses and glass-ceramics If a start- with the implant. This soft-tissue criti- ing implant composition is near the le cal isol, limit is shown by the dashed boundary, it may take only a few parts contour in Fig. 8, based on the work of per million of multivalent cations to Wilson and N~lletti.~' shrink the IB boundary to zero and elimi- The thickness of the bonding zone nate bioactivity Also, the sensitivity of between a bioactive implant and bone fit of a bioactive implant and length of is proportional to its Is value (compare time of immobiliztion postoperatively Fig. 1 with Fig. 2). Figure 13 shows an depends upon the Is value and close- optical micrograph of the interface of ness to the IB=O boundary Implants 45S5 Bioglasss bonded to bone (rat near the IB boundary require more tibia) after 1 year and an electron precise surgical fit and longer fixation microprobe analysis of the interface. times before loading The bioactive The bulk bioactive glass implant (BG), phases used in composites must have silica-rich layer (S), HCA layer (CA, P), Is values well away from the boundary and bone (B) are indicated. Living to start to bond within the 2 to 4 weeks bone cells (osteocytes labeled 0) are required clinically at the interface. There is no seam of Bioactive implants with intermediate interfacial fibrous tissue. Even at the Is values do not develop a stable soft- electron microscopic level there is al- tissue bond instead the fibrous inter- most no unmineralized tissue at the face progressively mineralizes to form bonding interface, as reviewed by bone Consequently, there is a critical Gross et aL5 and Hench and Clark.44 ISOI~boundary beyond which bioactivity The thickness of the bonding region is is restricted to stable bone bonding about 100 pm. Inside the critical isoIB boundary the In summary, the reaction stages that bioactivity includes both stable bone appear to be involved in forming the bond of bioactive glasses and bioactive glass-ceramics with tissues are depicted in Fig. 14. The time dependence of the stages shown at the left correspond to implants with high le values. When these stages are length- ened the Is values decrease drarnati- cally. If stages 4 and 5 are delayed too long the implant no longer is bioactive. The biological events that occur early in the bonding process are still being established, as discussed by Da~ies;'~,*~ however, the sequence of biological steps appears to be stages 6 to 11 as shown in Fig. 14.

Silicon-richACalcium-& Bone -

6 20 40 60 sb 1iO 1iO 140 Distance across interface (pn) Fig. 13. (a) Optical micrograph of a (BG) 4585 Bioglass@implant bonded to (B) rat bone after 1 year showing (0)osteocytes or bone cells in conjunction with the (Ca-P) HCA layer formed on top of the (S) silica gel (from Ref 44) (b) Electron microprobe analysis across the implant-bone interface shown in (a) (electron microprobe at 20 kV, 100 nA (speciman current), -1-pm beam diameter, and Fig. 14. Sequence of interfacial reactions involved in forming a 20 pm/(min in) scan rate) (from Ref 44) bond between tissue and bioactive ceramics July 1991 Biocerumics: From Concept fo Clinic 1499

The failure strength of a bioactively vents infection from migrating down ---*- fixed bond appears to be a nonlinear the electrodes thereby protecting the function of the IBvalue of the implant. patient, who must untiug the electron- Instead, interfacial strength appears to ics at night. Figure 15(a) shows the be inversely dependent on the thick- University of London extracochlear Table VI. Present Uses of ness of the bonding zone. For example, electrode array and its implantation Biocerarnics 45% Bioglasse with a very high site. Figure 15(b) shows the implant in lBvalue develops a gel-bonding layer place in a patient. Orthopedic load-bearing applications of 100 pm (Fig. 13) which has a rela- A1203 IX. Calcium Phosphate Ceramics Stabilized zirconia tively low shear strength. In contrast, PE-HA composite A/W glass-ceramic, with an intermedi- Calcium phosphate-based bioce- ate lBvalue has a bonding interface in ramics have been in use in medicine Coatings for chemical bonding the range of 10 to 20 pm and a very and dentistry for nearly 20 years, as (orthopedic, dental, and maxillofacial high resistance to shear (see Fig. 7). reviewed by Hulbert et a/.,3de Gr~ot,'~"~prosthetics) de Groot and Le Geros," Jar~ho,~~and HA Thus, the interfacial bonding strength Bioactive glasses appears to be optimum for IBvalues of Williams.8zApplications include dental Bioactive glass-ceramics -4. However, it is important to recog- implants, percutaneous devices, and nize that the interfacial area for bonding use in periodontal treatment, alveolar Dental implants is time dependent, as shown in Fig. 1. ridge augmentation, orthopedics, AI,O, Therefore, interfacial strength is time maxillofacial surgery, otolaryngology, HA ~ Bioactive glasses dependent and is a function of mor- and spinal surgery (Table VI). Differ- phological factors such as the change ent phases of calcium phosphate ce- Alveolar ridge augmentations in interfacial area with time, progres- ramics are used depending upon A1203 sive mineralization of the interfacial tis- whether a resorbable or bioactive ma- HA sues, and a resulting increase of the terial is desired. HA-autogenous bone composite elastic modulus of the interfacial bond The stable phases of calcium phos- HA-PLA composite Bioactive glasses as well as a function of shear strength phate ceramics depend considerably per unit of bonded area. Few data exist upon temperature and the presence of Otolaryngological to quantify these variables. water, either during processing or in AID, Clinical applications of bioactive the use environment. At body tempera- HA - glasses and glass-ceramics are ture only two calcium phosphates are Bioactive glasses reviewed by Gross et Yamamuro stable in contact with aqueous media, Bioactive glass-ceramics et and Hench and Wilson30 (see such as body fluids; at pH <4.2 the Artificial tendon and ligament Table VI). The 8-year history of stable phase is CaHPO, . 2H20 (dical- PLA-carbon-f iber composite successful use of CeravitaP glass- cium phosphaten or brushite, C2P), whereas at pH 24.2 the stable phase Artificial heart valves ceramics and 4595 Bioglass@implants Pyrolytic carbon coatings in middle-ear surgery to replace os- is Ca,0(P04)6(OH)2(HA). At higher sicles damaged by chronic infection, temperatures other phases, such as Coatings for tissue ingrowth (cardiovas- reported by Re~k,'~Merwin,86 and Ca,( P04)2(p-tricalcium phosphate, cular, orthopedic, dental, and rnaxillofa- Do~ek,'~are especially encouraging as C3P, or TCP) and Ca4Pz0, (tetracal- cia1 prosthetics) is the use of A/W glass-ceramic in re- cium phosphate, C,P) are present. The A1203 placing the iliac crest and in vertebral unhydrated high-temperature calcium Temporary bone space fillers surgery by Yamam~ro.~~~~~~'~45S5 Bio- phosphate phases interact with water, TCP glass@implants have been used sucess- or body fluids, at 37°C to form HA. The Calcium and phosphate salts fully for maintenance of the alveolar HA forms on exposed surfaces of TCP by the following reaction: Periodontal pocket obliteration ridge for denture wearers for up to HA 7 yearsg0 with nearly a 90% retention 4Ca3(P04)2(~)+2H20 HA-PLA composite rate, as reported by Stanley et a/." TCP Many patients are profoundly deaf -Ca,,(PO,),(OH),(surface) Calcium and phosphate salts because of loss of the auditory nerve Bioactive glasses +2Ca2++2HPOf- (2) fibers in the cochlea. Research has Maxillofacial reconstruction shown that such patients can still A1203 "hear" electronic impulses applied to Thus, the solubility of a TCP surface approaches the solubility of HA and HA the cochlea, although at much lower HA-PLA composite frequency discrimination. These find- decreases the pH of the solution which Bioactive glasses ings have led to the development of an further increases the solubility of TCP and enhances resorption. Williamse2 Percutaneous access devices extracochlear electrical implant which Bioactive glass-ceramics delivers electrical signals in real discusses the importance of the Ca/P ratio in determining solubility and ten- Bioactive glasses time encoded by a computer to match HA speech pattern^.^' A team led by dency for resorbtion in the body. The Fourcin and Douek in London reportg' presence of micropores in the sintered Orthopedic fixation devices successful clinical use of an array of material can increase the solubility of PLA-carbon fibers these phases." 13,33 PLA-calcium phosphate-based four platinum electrodes, insulated by glass fibers medical-grade alumina and anchored in Sintering of calcium phosphate ce- the bone with a 45S5 Bioglass@sleeve. ramics usually occurs in the range of Spinal surgery The alumina provides mechanical and 1000" to 1500°C following compaction Bioactive glass-ceramic dielectric stability to the device and the of the powder into a desired shape.33 HA bioactive glass bonds both to the bone The phases formed at high tempera- and the soft connective tissues as it ture depend not only on temperature protrudes through the skin. This her- but also on the partial pressure of metic, percutaneous living seal both water in the sintering atmosphere. provides mechanical stability and pre- 'Also called calcium monophosphate!' 1500 Journal of the American Ceramic Sociery - Hench Vol. 74, No. 7

With water present, HA can be formed yield a precipitate of stoichiometric and is a stable phase up to 136OoC, HA. The Ca", PO:-, and OH- ions as shown in the phase equilibrium can be replaced by other ions dur- diagram for CaO and P205with ing processing or in physiological 500-mmHg (-66-kPa) partial pressure surroundings; e.g., fluorapatite, of water (Fig 16) Without water, C4P Ca,,(P04)s(OH)2~,where OI00 ,urn in diameter, created to permit bone growth). The dependence of Compressive strength (a,) and total pore volume (V,) is described by de Groot et a/.'3as

u,= 700 exp( - 5V,) MPa (3) where V, is in the range 0 to 0.5. Tensile strength (u,)depends greatly on the volume fraction of microporosity (Vm).

at=220 exp(-20Vm) MPa (4) The Weibull factor (n) of HA implants is low in physiological solutions (n=12), which indicates low reliability under

a'C3P+ liquid 1700

1600

h 1500 2 c e$ I-" 1400

1300

1200 I I I 65 60 HA CjP 50 tCaO (wtO/o)

Fig. 15. (a) Extracochlear electrical implant for the profound deaf Fig. 16. Calcium phosphate phase equilibrium diagram with showing (center) electrode array (upper left) implantation site, and 500-mmHg (-66-kPa) partial pressure of water Shaded area is proc- (lower right) schematic of the components (b) Implant in patient's essing range to yield HA containing implants (Diagram based ear, with electronics disconnected (Photographs courtesy E upon Ref 13 ) Douek ) July 1991 Bioceramics: From Concept to Clinic 1501 tensile loads. Consequently, in clinical practice calcium phosphate bioceramics should be used as (1) powders, (2) small, unloaded implants such as in the middle ear, (3) dental implants with reinforcing metal posts, (4) coatings on metal implants, (5) low-loaded porous implants where bone growth acts as a reinforcing phase, or (6) the bioactive phase in a polymer-bioactive ceramic composite. The bonding mechanisms of dense HA implants appear to be very different from those described earlier for bioactive glasses. Evidence for the bonding process for HA implants is reviewed by Jar~ho.~~A cellular bone matrix from differentiated osteoblasts appears at the surface, producing a narrow amorphous electron-dense Fig. 17. TEM micrograph using lattice imaging to show epitaxial band only 3 to 5 pm wide. Between bonding between (bottom) HA implants and (top) the HA phase of this area and the cells, collagen bun- bone (Photograph courtesy of M Ogiso) dles are seen. Bone mineral crystals have been identified in this otherwise amorphous area.g‘ As the site matures, the bonding zone shrinks to a depth of only 0.05 to 0.2 pm. The result is normal bone attached through a thin epitaxial bonding layer to the bulk implant.” Ogiso et a/.9394have shown, through transmission electron microscopy (TEM) lattice image analysis of dense HA bone interfaces, an almost perfect epitaxial alignment of the growing bone crystallites with the apatite crystals in the implant (Fig. 17). A consequence of this ultrathin bonding zone is a very high gradient in elastic modulus at the bonding interface between HA and bone. This is one of the major differences between the bioactive apatites and the bioactive Fig. 18. Stable HA implant in forearm of a volunteer 6 years after glasses and glass-ceramics, as illus- implantation. (Photograph courtesy of H. Aoki.) trated in Fig. 2. The implications of this difference on the interfacial response of that a dense, sintered HA implant, used the implant to applied stress and bone in contact with skin and subcutaneous vitality is discussed in Ref. 1 (Ch. 14). tissues, provides a stable seal and has The significance of this difference clini- prevented downgrowth of epidermis cally is as yet unknown. and infiltration of cells, common char- Potentially, one of the most impor- acteristics of glassy carbon or silicone tant applications of sintered, dense HA rubber percutaneous leads. Figure 18 implants is as percutaneous access shows a HA implant in the arm of a assist devices (PAD)95for continuous volunteer after 6 years without any ambulatory peritoneal dialysis (CAPD) serious infection. X-ray computer patients. There are hundreds of thou- tomography of the forearm showed no sands of patients worldwide who calcification of the soft tissues around undergo renal dialysis and could bene- the skin button.97 fit from CAPD, which would decrease Two patients suffering from end- their dependency on the resources re- stage renal disease developed arterio- quired for conventional hemodialysi~.~~venous fistulas after 10 months and 12 However, a severe limitation on CAPD years hemodialysis, and therefore, is the incidence of peritonitis, up to could not continue treatment. After im- 1.5 incidents per year per patient, usu- plantation of an HA PAD and CAPD ally caused by invasion of bacteria treatment, they have remained well for along the wall of the inserted peri- more than 2 years,96 indicating the toneal catheter. Catheters used, now value of this application of bioactive made of silicone rubber, do not form a ceramics. completely sealed junction with the skin or subcutaneous tissues and X. Composites thereby provide a pathway for infection One of the primary restrictions on to occur. clinical use of bioceramics is the uncer- Aoki and c011eagues~~~~~have shown tain lifetime under the complex stress 1502 Journal of the American Ceramic Society - Hench Vol. 14, No. I

states, slow crack growth, and cyclic cases the goal is to increase flexural fatigue that arise in many clinical appli- strength and strain to failure and de- Table VII. Number of cations. Two creative approaches to crease elastic modulus. Centers Represented at these mechanical limitations are use of Bonfield'"' argues that analogous im- World Biomaterials Congress bioactive ceramics as coatings or in plant materials with similar mechanical Year composites. Much of the rapid growth properties should be the goal when in the field of bioactive ceramics, bone is to be replaced Because of the Country 1980 1984 1988 summarized in Table VII is due to anisotropic deformation and fracture United States 267 development of various composite and characteristics of cortical bone, which FRG 212 coating systems, as reviewed by Doyle,99 is itself a composite of compliant colla- Netherlands 112 gen fibrils and brittle HCA crystals, the Japan 1 3 14 Ducheyne and McGuckin,'" and Italy 12 Soletz."' Many papers in Refs. 25 and Young's modulus (E) varies between -7 United Kingdom 2 41 also discuss these topics. Tables Vlll to 25 GPa, the critical stress intensity France 3 and IX show that composites and ranges from -2 to 12 MPa'm''', and Finland 1 coatings involve all three types of the critical strain intensity increases Belgium 1 biomaterials-nearly inert, resorbable, from as low as -600 J.m-2 to as DRG 1 and bioactive-and reflect differing ra- much as 5000 J.m-2, depending on Total 6 12 35 tionales in their development. In most orientation, age, and test c~ndition.'~~ In contrast, most bioceramics are much stiffer than bone and many ex- Table VIII. Bioceramic Composites hibit poor fracture toughness (Table Ill). Matrix Reinforcing phase* Reference Consequently, one approach to achieve Type 1 Nearly inert composites properites analogous to bone is to Polysufone Carbon fiber 99, 101 stiffen a compliant biocompatible syn- Polyethylene Carbon fiber 99 101 thetic polymer, such as PE, with a Poly(methy1 methacrylate) Carbon fiber higher modulus ceramic second phase, Carbon Carbon fiber 99 such as HA powder."' The effect is Carbon sic 101 Epoxy resin Alumina/stainless steel 101 to increase Young's modulus from 1 to 8 GPa and to decrease the strain to Type 2 Porous ingrowth composites failure from >90% to 3% (Fig. 19) as Coral HA goniopora DL polylactic acid 105 the volume fraction of HA increases to Type 3 Bioactive composites 0.5. The transition from ductile to brittle Bioglassa Stainless-steel fibers 100, 101 behavior occurs between 0.4 and 0.45 Bioglassm Titanium fiber 100, 101 volume fraction HA. Bonfield"" reported Collagen HA 99, 101 that the ultimate tensile strength of the Polyethylene HA 102 101 Poly(methy1 methacrylate) Phosphate-Silicate-Apatite glass fiber 99, 101 composite remained within the range of Polymer Phosphate glass 99 101 22 to 26 MPa. At 0.45 volume fraction HA/collagen Gelat in-resorcinol-formaldehyde 99 HA, the Klc value was 2.920.3 MPa A/W glass-ceramic Transformation-toughened Zr02 106 m1'2, whereas at <0.4 volume fraction HA the fracture toughness was consid- Type 4 Resorbable composites PLA/PGA PLA/PGA fibers 99, 101 erably greater, because of the ductile Pol y hyd roxy but urate HA 99, 101 deformation associated with crack PLA/PGA HA propagation. Thus, the mechanical 99.I~ 101 *DL IS dextra levorotatory (left handed) properties of the PE-HA composite are close to or superior to those of bone. The bioactive phase is exposed by Table IX. Bioceramic Coatings machining the surface of the com- Substrate Coating Reference posite. Implant tests of the PE+0.4HA 316L stainless steel Pyrolytic carbon Boc kros 'lo composites have demonstrated devel- 316L stainless steel 4585 BioglassQ Hen~h'~ opment of bone bonding between the 316L stainless steel aAl2O3-HA-TiN Hayashi" natural hard tissue and the synthetic 316L stainless steel HA Munt ing25 implant.'04An application of the PE-HA 316L stainless steel SE51(45S5) Bioglasse (to4' composite currently in test in London Co-Cr alloy 4555 Bioglasso and Lacef ield is as a medullary sleeve for fixation of 52S4.6 Bioglassm the stem of femoral head replacements, Co-Cr alloy HA Munting'5 where the bonding of the composite to Ti-6AI-4V alloy 45% Bioglasso West4' bone should prevent stress shielding Ti-6AI-4V alloy HA/ABS Glass* Ma runs '' because of its match of Young's modu- Ti-6AI-4V alloy HA Ducheyne" Ius with the bone. Figure 20 depicts Ti-GAI-4V alloy TCP Cook114, such a device. Lacefield", Another promising approach toward Chae" achieving high toughness, ductility, and Ti-6AI-4V alloy HA Hamada'' a Young's modulus matching that of Ti-6AI-4V alloy A1203 Hamada" bone was developed by Ducheyne and Ti-6AI-4V alloy+porous beads HA Oon i s hi 25 the author.lo7This composite uses sin- Ti-6AI-4V alloy TiOz-HA Hayashi25 Ti-6AI-4V alloy HA, TRP, TCP de Gr~ot*~ tered 316L stainless steel of 50-, 100-, Ti-6A1-4V alloy+Ti powder HA DucheyneZ5 or 200-~mdiameter or titanium fibers, Ti-6AI-4V alloy HA Kay" which provide an interconnected fibrous matrix which is then impregnated with Alumina 4585 BioglassQ Green~pan~~ molten 4585 Bioglass@'.After the com- *ABS is alkali borosilicate glass posite is cooled and annealed, a very strong and tough material results, with July 1991 Bioceramics: From Concept to Clinic 1503

E (cortical bone) 100

0 I I I I 0.1 0.2 0.3 0.4 Volume fraction HA Fig. 19. Effect of HA on Young’s modulus and strain to failure of a Fig. 20. PE-HA composite used as a sleeve for the stem of a PE-HA composite (Based upon data of Ref 102) total hip implant (Photograph courtesy of W Bonfield) metal-to-glass volume ratios between heart surgery soon thereafter.”’ The 416 to 614. Strength enhancements of first time the low-temperature isotropic up to 340 MPa are obtained in bending (LTI) carbon coatings were used in hu- with substantial ductility of up to 10% mans was as a prosthetic heart valve elongation, as shown by the set of by DeBakey in 1969.”’ bend-test bars in Fig. 21, which bent Almost all commonly used prosthetic 90” without fracturing. When an outer heart valves today have LTI carbon coating of BiogIassB is maintained, coatings for the orifice and/or occluder the composite implants bond to bone (Fig. 22) because of their excellent tissue and perform a load-carrying resistance to blood clot formation function.lo8 and long fatigue life.”’ More than The strongest of the bioactive ce- 600 000 lives have been prolonged ramic composites is composed of A/W through the use of this bioceramic in glass-ceramic containing a dispersion heart valves. of tetragonal zirconia, developed by Three types of carbon are used in Kasuga et a/.’06 Volume fractions of biomedical devices: the LTI variety of 0.5 zirconia lead to bend strength val- pyrolytic carbon, glassy (vitreous) car- ues of 703 MPa and Klc values of bon, and the ultralow-temperature 4 MPa and also formed HA on isotropic (ULTI) form of vapor- their surface under in vitro test condi- deposited ~arbon.”‘~”~These carbon tions. These results, with other com- materials are integral and monolithic posite systems listed in Table Vlll, materials (glassy carbon and LTI car- indicate the great potential of bioactive bon) or impermeable thin coatings and resorbable bioceramics with good (ULTI carbon). These three forms do mechanical behavior for a wide variety not suffer from the integrity problems of clinical applications. As yet, how- typical of other available carbon ma- ever, very little data exist on the envi- terials. With the exception of the LTI ronmental sensitivity and fatigue life of carbons codeposited with silicon, all these composite systems under physi- the carbon materials in clinical use are ological loads and environments. Until pure elemental carbon. Up to 20 wt% silicon has been added to LTI carbon these data are acquired caution must Fig. 21. 4585 Bioglass@composite rein- be excercised when using such materi- without significantly affecting the bio- forced with 60% of 100-pm 316L stainless- als climcal\y, other than in high\y con- compatibility of the material. The com- steel fibers tested in bending Note a be;: trolled trials. position, structure, and fabrication of angle of >90” is obtained without failure the three clinically relevant carbons are (Photograph courtesy of P Ducheyne) unique when compared with the more XI. Coatings common, naturally occurring form of (1) Carbon carbon (i.e., graphite) and other indus- Bokrosiog applied, in 1967, for a trial forms produced from pure elemen- patent describing the medical use of tal carbon. pyrolytic carbon coatings on metal The LTI, ULTI, and glassy carbons substrates. The coatings were used in are subcrystalline forms and represent 1504 Journal of the American Ceramic Society - Hench Vol. 14, No. 7

concentrated or point loading without galling or incurring surface damage. The bond strength of the ULTl carbon to stainless steel and to Ti-6AI-4V ex- ceeds 70 MPa as measured with a thin-film adhesion tester. This excellent bond is, in part, achieved through the formation of interfacial carbides. The ULTl carbon coating generally has a lower bond strength with materials that do not form carbides. Another unique characteristic of the turbostratic carbons is that they do not fail in fatigue. The ultimate strength of turbostratic carbon, as opposed to metals, does not degrade with cyclical loading. The fatigue strength of these carbon structures is equal to the single- cycle fracture strength. It appears that, unlike other crystalline solids, these Fig. 22. LIT pyrolitic carbon-c:oated heart valves. (Photograph forms of carbon do not contain mobile courtesy of J. Bokros.) defects, which at normal temperatures can move and provide a mechansim for the initiation of a fatigue crack. a lower degree of crystal perfection. The more important known proper- There is no order between the layers ties of the turbostratic carbons are such as there is in graphite; therefore, listed in Ref. 113. the crystal structure of these carbons Carbon surfaces are not only is two dimensional. Such a structure, thromboresistant, but also appear to called turbostratic, has densities be- be compatible with the cellular ele- tween about 1400 and 2100 kg.m-3. ments of blood; they do not influence High-density LTI carbons are the plasma proteins or alter the activity of strongest bulk forms of carbon and plasma enzymes. One of the proposed their strength can further be increased explanations for the blood compatibil- by adding silicon. ULTl carbon can also ity of these materials is that they ad- be produced with high densities and sorb blood proteins on their surface strengths, but it is available only as a without altering them. thin coating (0.1 to 1.0 pm) of pure Reference 113 summarizes the uses carbon. Glassy carbon is inherently a of glassy, LTI, and ULTl carbons in low-density material and, as such, is various medical areas. weak. Its strength cannot be increased through processing. Processing of all (2) Hydroxyapatite three types of medical carbons are discussed in Ref. 113 (Haubold et a/.). A second bioceramic coating which The mechanical properties of the has reached a significant level of clini- various carbons are intimately related cal application is the use of HA as to their microstructures. In an isotropic a coating on porous metal surfaces carbon, it is possible to generate ma- for fixation of orthopedic prostheses. terials with low elastic moduli (20 GPa) This approach combines types 2 and and high flexural strength (275 to 3 methods of fixation (Table II) and 620 MPa). There are many benefits as originates from the observations of a result of this combination of proper- Ducheyne and colleague^,^ in 1980, ties; e.g., large strains (-2%) are pos- that HA powder in the pores of a sible without fracture. The turbostratic porous, coated-metal implant would carbons are very tough when compared significantly affect the rate and vitality with ceramics such as aluminum oxide. of bone ingrowth into the pores. A large The energy to fracture for LTI carbon number of investigators have explored is approximately 5.5 MJ.m-3, com- various means of applying the HA coat- pared with 0.18 MJ.m-3 for aluminum ing, as discussed in Refs. 3, 5, 12, 25, oxide; i.e., the carbon is approximately 41, and 101, with plasma spray coating 25 times as tough. The strain to frac- generally being preferred. There is a ture for the vapor-deposited carbons is substantial enhancement of the early greater than 5.0%,making it feasible to stage interfacial bond strength of im- coat highly flexible polymeric materials plants with a plasma-sprayed HA coat- such as PE, polyesters, and nylon with- ing when compared with porous metals out fear of fracturing the coating when without the coating, as illustrated in the substrate is flexed. Fig. 23, from C00k.”~However, long- term animal studies and clinical trials The turbostractic carbon materials of load-bearing dental and orthopedic have extremely good wear resistance, prostheses suggest that the HA coat- some of which can be attributed to ings may degrade or come off, and the their toughness, i,e., their capacity to clinical consequences are still being sustain large local elastic strains under debated ,84 July 1991 Bioceramics: From Concept to Clinic 1505 XII. Resorbable Calcium Phosphates Resorption or biodegradation of calcium phosphate ceramics is caused by (1) physiochemical dissolution, which depends on the solubility prod- uct of the material and local pH of its environment (New surface phases may be formed, e.g., amorphous calcium phosphate, dicalcium phosphate dihydrate, octacalcium phosphate (Ca4(P04),H 3H20),and anionic substi- tuted HA ); (2) physical disintegration into small particles due to preferential chemical attack of grain boundaries; and (3) biological factors, such as phagocytosis, which causes a decrease in local pH.'* All calcium phosphate ceramics I I I I I I 1 biodegrade to varying degrees in the I 01234567 following order: a-TCP>p-TCP> >HA. Strength (MPa) The rate of biodegradation increases as (1) surface area increases (pow- Uncoated ders>porous solid>dense solid), HA coated (2) crystallinity decreases, (3) crystal perfection decreases, (4) crystal and Fig. 23. Comparison of interfacial bond strength of porous titanium with and without plasma-sprayed HA coatings (Photograph (5) grain size decrease, and ionic sub- courtesy of S Cook ) stitutions of CO,'-, Mg", Srz+ in HA take place. Factors which result in a decreasing rate of biodegradation in- be delivered in this manner, whereas a clude (1) F-substitution in HA, (2) Mg2+ maximum of 3000 rd of external radia- substitution in 0-TCP, and (3) decreas- tion can be tolerated by the patient. In ing p-TCP/HA ratios in biphasic cal- a clinical trial in Canada, Boos et a/."5 cium phosphates. Because of these report response for 35 of 46 patients variables it is necessary to control the with carcinoma of the liver: complete microstructure and phase state of a remission of 1, partial remission of 6, resorbable calcium phosphate bioce- and stability of the disease of 24, with ramic in addition to achieving precise a mean survival of 16.1 months for the compositional control to produce a responders versus only 8.8 months for given rate of resorption in the body. As nonresponders. yet, there are few data on the kinetics Dayi16 reports that the glass- of these reactions and the variables in- microsphere radiation delivery vehicle fluencing the kinetics. can be modified with different radioactive isotopes to achieve various XIII. Therapeutic Applications ranges and is being tested preclinically A significant problem in the radiation for treatment of kidney cancer and treatment of cancer is the serious sys- arthritis. temic side effects. Localization of the Another approach to cancer treat- radiation at the site of the tumor de- ment using bioceramics is under creases the radiation dosage required development by Yamamuro in Kyoto, to kill the cancer cells and thereby Japan."' Ohura et have incorpo- minimizes side-effect toxicities. An in- rated 40 wt% Fe20, in a bioactive novative approach to the localized de- glass-ceramic composed of CaO- livery of radioactive yttrium-90 ("Y) to Si02-B,0,-P205. The magnetite phase treat liver cancer has been developed produces a high-saturation magnetiza- by Day1l4at the University of Missouri- tion which can be used to generate lo- Rolla using glass microspheres. A calized heating at an implant site. yttria-aluminosilicate glass, containing Temperatures of >42"C can be gener- *'Y is made in the form of 25-pm ated in bone, sufficient to kill tumor microspheres. Prior to use in hepatic cells, by applying a 300-0e, 100-Hz arterial infusion therapy, the micro- magnetic field. The wollastonite phase spheres are bombarded by neutrons which grows during crystallization pro- which creates 'OY, a radioactive isotope vides bonding of the granules to bone. which is a short half-life (64 h), short- Preclinical tests are underway. range (2.5 to 3 mm in the liver) p emit- Still another therapeutic application ter. The microspheres are injected of bioceramics is delivery of various through a catheter placed in an artery, steroid hormones from aluminum cal- and the blood stream carries them to cium phosphate porous the liver where a high proportion goes The advantage of this method is sus- to the cancerous part because of the tained delivery of a potentially toxic -3 times increased blood supply. A lo- substance over long periods of time, calized dosage of up to 15000 rd can again inhibiting systemic side effects 1.506 Journal of the American Ceramic Society - Henc:h Vol. 74, No. 7 due to large dosages. Details of the is proved before clinical use is ex- processing, delivery rates, and many panded. chemical and biological tests are given by Benghussi and Bajpai."' Acknowledgments: The author acknowl- edges the long-term collaboration with June XIV. Summary Wilson-Hench, wife and coresearcher, who helped in proofing: Alice Holt for manuscript Bioceramics has evolved to become preparation; and Professor Werner Lutze for a an integral and vital segment of our critical review. modern health-care delivery system. The full potential is only beginning to References be recognized. In the years to come ' L. L Hench and E C Ethridge, Biomaterials An the composition, microstructure, and hterfacral Approach. Academic Press, New York. molecular surface chemistry of various 1982 types of bioceramics will be tailored 'J. D Preston, "Properties in Dental Ceramics", in Proceedings of the fVth lnternationai Symposium to match the specific biological and on Dentai Maferials. Quintessa Publishing, metabolic requirements of tissues or Chicago, IL, 1988. disease states. This "molecular-based 3S.F. Hulbert, J C. Bokros, L L Hench, J pharmaceutical" approach to the de- Wilson, and G. Heimke, "Ceramics in Clinical Ap- sign of bioceramics should couple with plications Past, Present, and Future'' pp 189-213 the growth of genetic engineering, sen- in High Tech Ceramics. Edited by P. Vincenzim sor technology, and information proc- Elsevier, Amsterdam, Netheriands, 1987 essing, resulting in a range of products 4L L Hench, "Bioactive Ceramics", p. 54 in 610- and applications not even imagined at ceramics' Materiais Characteristics Versus In Vivo Behavior. Vol. 523 Edited by P Ducheyne and J present, but potentially beneficial to Lemons Annals of New York Academy of Sci- millions of people annually. ences, New York, 1988. However, at present there is a criti- 5U. Gross, R Kinne. H.J Schmitr, and V cal need for standard test methods to Strunz, "The Response of Bone to Surface Active determine the long-term lifetime per- Glass/Glass-Ceramics," CRC Crit. Rev Biocom- formance of bioceramics under realistic pa;, 4, 2 (1988) physiological loads. Too many materials P Christel, A. Meunier, J. M Dorlot, J.M Crolet. are introduced into clinical use without J Witvolet, L Sedel, and P.M Boutin, "Biome- proof tests or lifetime prediction tests. chanical Compatibility and Design of Ceramic Im- plants for Orthopedic Surgery"; p. 234 in Seldom are interinstitutional studies Bioceramics- Material Characteristics Versus In conducted to determine the reliability Vivo Behavior. Vol 523. Edited by P Ducheyne of biomechanical performance. Stand- and J Lemons Annals of New York Academy of ards do not presently exist for bio- Sciences, New York, 1988. active ceramics or coatings, although 'L L Hench, "Cementless Fixation", p 23 in the American Society for Testing and Biomateriais and Ciinical Appbations. Edited by Materials (ASTM) F-4 Committee on A Pizzoferrato, P G Marchetti, A Ravaglioli, and Medical and Surgical Materials and A J C Lee Elsevier, Amsterdam, Netherlands, Devices has a Subcommittee on 1987 Ceramic Materials (J. Lemons, Chair) 'J. Black, "Systemic Effects of Biomateriais," Riomater/ais, 5, 11 (1984) with a goal of achieving standards for 'P. Ducheyne, L L. Hench. A Kagan, M bioactive glasses and glass-ceramics Martens, A Burssens, and J.C Mulier. "The Effect ( D. Greens pan), h y d r ox y a pat it e of Hydroxyapatite Impregnation of Skeletal Bond- (H. Hahn), single-crystal and polycrys- ing of Porous Coated Implants," J. 6iomed. Mater tallhe alumina (H. Jung), and ceramic Res., 14, 225 (1980) coatings (J. Kay). Standardization of 'OK de Groot, Bioceramics of Calcium Phos- test methods also needs to be estab- phate, pp 1-32 CRC Press, Boca Raton, FL, lished. At present, there is no way to 1983 compare interfacial bonding strength "K de Groot, "Effect of Porosity and Physico- chemicai Properties on the Stabiiity, Resorption, data for the materials listed in Tables IV, and Strength of Calcium Phosphate Ceramics"; in VIII, or IX. Likewise, there is no basis to Bioceramics' Material Characteristics Versus In compare fatigue life for any of these Vivo Behavior, Vol. 523 Edited by P Ducheyne materials or establishing the relative and J Lemons Annals of New York Academy of importance of grain-boundary attack Sciences, New York, 1988 or slow crack growth under standard- "K de Groot and R Le Geros, "Significance of ized conditions We must correct these Porosity and Physical Chemistry of Calcium Phos- deficiencies within this decade because phate Ceramics", pp 268-77 in Bioceramics Mate- there is a rapidly growing number of ria/ Characteristics Versus In Vivo Behavior, Vol 523. Edited by P Ducheyne and J Lemons An- failures of load-bearing metallic pros- nals of New York Academy of Sciences, New York. theses with poly(methy1 methacrylate) 1988 (PMMA) cement fixation. The lifetime of I3K de Groot. C P.A.T Klein, J G C Wolke, cemented devices is often less than and J. de Blieck-Hogervorst, "Chemistry of Cal- the lifetime of a patient. Therefore, revi- cium Phosphate Bioceramics", pp 3-15 in Hand- sion surgery is required with great book of Bioactive Ceramics, Vol. 11, Calcium trauma to the patient and great Phosphate and Hydroxylapatite Ceramics. Edited expense to our health-care system. by T Yamamuro. L L Hench. and J Wilson. CRC Bioceramics offer one of the few alter- Press, Boca Raton, FL, 1990 "L L Hench, R J. Splinter, W C Alien, and T K natives for solving this problem. How- Greeniee, Jr , "Bonding Mechanisms at the Inter- ever, it is our responsibility to assure face of Ceramic Prosthetic Materials," J Riomed. that long-term reliability of bioceramics Mater Res., 2 [I1 117-41 (1972). July 1991 Bioceramics: From Concept to Clinic 1507

"P Griss and G Heimke, "Five-Years Experi- Bone Tissue." J Biomed Mater Res, 19, 685 ence with Ceramic-Metal Composite Hip Endo- (1985) prostheses " Arch Orthop Traumat Surg., 98, 35J Wilson, G H Pigott F J Schoen. and L L 157-65 (1981) Hench, "Toxicology and Biocompatibility of Bio- 'bP M Boutin "A View of 15-Years Results Using glass,'' J Biomed Mafer. Res , 15, 805 (1981) the Alumina-Alumina Hip Joint Protheses", p 297 36 T Kokubo, S Ito. S Sakka, and T Yamamuro, in Ceramics in Cbnicai Appiications. Edited by P "Formation of a High-Strength Bioactive Glass- Vincenzini Elsevier New York, 1987 Ceramic in the System MgO-Ca0-Si0,-P20,," "H Oonisihi, N. Okabe. T Hamaguchi. and T J Mater Sci, 21, 536 (1986). 37 Nabeshima, Orthopedic Ceramic impiants, Vol 1, T Kitsugi, T Yamamuro. and T Kokubo, p 157 Japanese Society of Orthopaedic Ceramic "Bonding Behavior of a Glass-Ceramic Containing Implants, Osaka, Japan, 1981 Apatite and Wollastonite in Segmental Replace- '*H Plenk, "Biocompatibility of Ceramics in ment of the Rabbit Tibia Under Load-Bearing Con- Joint Prostheses", pp. 269-95 in Biocompatibiiity ditions," J Bone Jt Surg., Am Voi, 71A, 264 of Orthopedic impiants. Vol 1 CRC Press, Boca (1989) 30 Raton, FL, 1982 S Yoshii, Y Kakutani. T Yamamuro, T "R V McKinney, Jr , and J Lemons, The Dentai Nakamura, T Kitsugi, M Oka. T Kokubo, and M implant, pp 31-50. PSG Publishing, Littleton, MA, Takagi. "Strength of Bonding Between A-W Glass- 1985 Ceramic and the Surface of Bone Cortex," J 20 E. Dorre and W Dawihl, "Ceramic Hip Endo- Biomed Mater Res., 22 [A] 327 (1988) prostheses ', pp 113-27 in Mechanical Properties '"T Kokubo, M Shigematsu. Y Nagashima, of Biomateriais. Edited by G W. Hastings and D F M Tashiro T Nakamura T Yamamuro. and S. Williams Wiley, New York, 1980 Higashi, "Apatite- and Wollastonite-Containing "J E Ritter, D C Greenspan R A Palmer, and Glass-Ceramics for Prosthetic Application," Buii L L Hench, "Use of Fracture Mechanics Theory in lnst Chem Res., Kyoto Univ, 60, 260 (1982) 40 Lifetime Predictions for Alumina and Bioglass- T Yamamuro, J Shikata. Y Kakutani S Coated Alumina," J. Biomed Mater. Res , 13, 251- Yoshii, T Kitsugi, and K Ono, "Novel Methods for 63 (1979) Clinical Applicatioins of Bioactive Ceramics", "R T Chiroff. E W White, J N Webber. and D p 107 in Bioceramics Materiai Characteristics Ver- Roy, "Tissue lngrowth of Replamineform Implants," sus In Vivo Behavior, Vol 523. Edited by P J Biomed. Mater Res Symp, 6, 29-45 (1975) Ducheync and J E Lemons Annals of New York i3E.W. White, J.N Webber, D M. Roy, E L Academy of Sciences. New York. 1988 Owen, R T Chiroff, and R. A White, "Replamine- "Handbook on Bioactive Ceramics, Vol I form Porous Biomaterials for Hard Tissue Implant Bioactive Glasses and Glass-Ceramics Edited by

Applications " J. Biomed. Mater. Res. Symp , 6. T Yamamuro, L L Hench, and J Wilson CRC 23-27 (1975) Press, Boca Raton. FL, 1990. 24R E Holmes. R W Mooney, R W Bucholr, and "J, Wilson and D Nolletti, "Bonding of Soft Tis- A. F. Tencer, "A Coralline Hydroxyapatite Bone sues to Bioglass'", pp 283-302 in Handbook of Graft Substitute," Ciin Orfhop. Reiat Res, 188, Bioactive Ceramics, Vol I, Bioactive Glasses and 282-92 (1984) Glass-Ceramics Edited by T Yamamuro, L L 75Handbook of Bioactive Ceramics, Vol II Cal- Hench, and J Wilson CRC Press, Boca Raton cium Phosphate and Hydroxylapatite Ceramics FL, 1990 Edited by T. Yamamuro, L L Hench, and J 43LL Hench and H A Paschall, "Direct Chemi- Wilson CRC Press, Boca Raton, FL, 1990 cal Bonding Between Bioactive Glass-Ceramic "S F Hulbert and J J Klawitter, "Application Materials and Bone." J Biomed Mater. Res. of Porous Ceramics for the Attachment of Symp, 4. 25-42 (1973) Load-Bearing Internal Orthopedic Applications." 4'L L Hench and A E Clark "Adhesion to Biomed Mafer Symp., 2 161-229 (1972) Bone", Ch 6 in Biocompatibiiity of Orfhopedic impiants, Vol. 2. Edited by D F Williams CRC Press "S IHulbert S J Morrison and J J Klawitter Boc?, Raton FL, 1982 'Tissue Reaction to Three Ceramics of Porous 45 0.H Andersson, G Liu, K.H Karlsson. L and Non-porous Structures,' J. Biorned Mater Niemi, J Miettincn and J Juhanoja, 'In Vivo Be- Res, 6, 347-74 (1972) havior of Glasses in the Si07-Na20-CaO-P20,- "S IHulbert. F W Cooke. J J Klawitter, R B AI,O,,-B,Oj System," J. Mater Sci., Mater Med., Leonard, B W Sauer, D D Moyle, and H.B 219-27 (1990) Skinner, "Attachment of Prostheses to the 46DC Greenspan and L L Hench, "Chemical Musculo-Skeletal System by Tissue lngrowth and and Mechanical Behavior of Bioglass-Coated Alu- Mechanical Interlocking," Biomed Mafer Symp., mina.' J Biomed Mater. Res., 10 [4] 503-509 4. 1-23 (1973). (1976). 29S.F. Hulbert. J R Matthews. J J Klawitter, "A W Ham. Histoiogy, 7th ed; 378-433 Lippin- B W Sauer, and R B Leonard, "Effect of Stress cott. Philadelphia, PA, 1969 on Tissue lngrowth into Porous Aluminum Oxide," 40 G W Hastings and P Ducheyne eds , Natural Biomed. Mafer Symp., 5, 85-97 (1974) and Living Biomateriais, 27-98. CRC Press, Boca Raton, FL 1984. "L L Hench and J.W Wilson. "Surface-Active 49 M Gross and V Strunz, "The Anchoring 01 Biomaterials.' Science (Washingtoii, D C), 226. U Glass Ceramics of Different Solubility in the Femur 630 (1984) of the Rat," J, Biomed Mater Res, 14, 607 (1980) 3'U Gross and V Strunz, "The Interface of Vari- "U Gross, J Brandes, V. Strunz. J Bab. and J ous Glasses and Glass-Ceramics with a Bony Im- Sela, "The Ultrastructure of the Interface Between plantation Bed." J Biomed. Mater. Res, 19, 251 a Glass Ceramic and Bone," J. Biomed Mater (1985) 32W Hohland. W Vogel, K Naurnann, and J Res, 15, 291 (1981) 5'U Gross, H J Schmitz, V Strunz, D Gummel, "Interface Reactions Between Machin- Schuppan, and J Termine, "Proteins at the lnter- able Bioactive Glass-Ceramics and Bone," face of Bone-Bonding and Non-Bonding glass^ J Biomed. Mafer Res , 19. 303 (1985) Ceramics". p 98 in Transactions the Twelfth 33M,Jarcho, "Calcium Phosphate Ceramics as of Annual Meeting of the Society for Biomaterials Hard Tissue Prosthctics " Ciin Orthop Rekt Society for Biomaterials. Algonquin. iL 1986 Res , 157, 259 (1981) 34T Nakamura, T Yamamuro, S Higashi. T 52uGross, W Roggendorf, H. J Schmitz. and Kokubo and S Itoo, "A New Glass-Ceramic for V Strunz, "Testing Procedures for Surface Reac- Bone Replacement Evaluation of its Bonding to tive Biomaterials' p 367 in Bioiogicai and Biome- Vol. No. 1508 Journal of the American Cerumic Society - Hench 74, 7

chanicai Performance of Biomaterials Edited by P cal Engineering, ibo Akademi, Finland, 1990. Christel, A Meunier, and A J C Lee Elsevier, Am- "CY Kim, A E Clark, and L L Hench, "Early sterdam, Netherlands, 1986 Stages of Calcium Phosphate Layer Formation in 53W Vogel, W Hohland, K Naumann. J Vogel, Bioglasses@,"J Non-Cryst Soiids, 113. 195-202 G Carl, W Gotr, and P Wange. "Glass-Ceramics (1989) tor Medicine and Dentistry", p 353 in Handbook "0 H Andersson and K H. Karlsson, "On the of Bioactive Ceramics, Vol I, Bioactive Glasses Bioactivity of Sillcate Glass," J, Non-Cryst Soiids, and Glass-Ceramics. Edited by T Yamamuro, L. L. 129. 145-51 (1991) Hench, and J Wilson CRC Press, Boca Raton, 73D M Sanders, W.B Person, and L L Hench FL, 1990 "Quantitative Analysis of Glass Structure Using In- "'0 H Andersson. K H Karlsson. K frared Reflection Spectra." Appl Spectrosc , 28 Kangasniemi, and A Yti-Urpo, "Models for Physi- [3]247--55 (1974). cal Properties and Bioactivity of Phosphate Opal 74B0 Fowler, "Infrared Studies of Apatites I. Glasses," Giasfech. Ber,, 61, 300-305 (1988) Vibrational Assignments for Calcium, Strontium. 55K Kangasnieini and A Yti-Urpo. "Biological and Barium Hydroxyapatite Utilizing Isotopic Sub- Response to Glasses in the Si02-Na10-Ca0- stitutioii," inorg Chem, 13 [l] 194 (1974) P205-B203System'. pp 97-108 in Handbook of "R.F Le Geros. G Bone, and R Le Geros. Bioactive Ceramics, Vol I, Bioactive Glasses and "Type of H20 in Human Enamel and in Precipi- Glass~Ceramics,Edited by T Yamamuro. L L tated Apatites." Caici! Tissue Res, 26, 111 (1978) Hench. and J Wilson CRC Press, Boca Raton, "1 L Hench, 'Stability of Ceramics in the FL. 1990 Physiological Environment", pp 67-85 in Funda- "L L Hench, D B Spilmaii and D E Nolletti, mentai Aspects of Biocompalibiiifx Vol 1 Edited Fluoride Bioglasses@)",pp 99-104 in Biological by D F Williams CRC Press, Boca Raton FL, and Biomechanicai Performance of Biomateriais 1981 Edited by P Christel, A Meunier, and A J C Lee ''T Kokubo, "Bonding Mechanism of Bioactive Elsevier, Amsterdam, Netherlands 1986 Glass-Ceramic A-W to Living Bone", pp 41-50 in "L L Hench, H A Paschall, W C Allen, and G Handbook of Bioactive Ceramics, Vol I, Bioactive Piotrowski. "Interfacial Behavior of Ceramic Im- Glasses and Glass-Ceramics Edited by T plants," Nati. Bur Stand. Spec Pub/, 415, 19-35 Yamamuro, L i.Hench and J Wilson. CRC Press, (1975) Boca Raton, FL, 1990 "L L Hench and H A. Paschall, "Histo- '"C Oktsuk!, T Kokubo. K Takatsuka, and T Cheinical Responses at a Biomaterials Interface," Yarnamuro, "Compositional Dependence of Bioac- J Biomed Mater. Res , 5 [I] 49-64 (1974) tivity of Glasses in the System CaO-SiOP-PiOs "W J McCrackeii, D.E Clark and L L Hench, and its In Vitro Evaluation," Nippon Seramlkkusu "Aqueous Durability of Lithium Disilicate Glass- Kyohai Gakujutsu Ronbunshi, 99 [l] 1-6 (1991) Ceramics," Am Ceram. SOC Bull., 61 1111 1218- "R LI, A E Clark and L L Hench, 'An Investi- 23 (1982) gation of Bioactive Glass Powders by the Sol-Gel "D E Clark, C G Pantano, and L L. Hench, Process", p 40 in Transactions of 16th Annual Corrosion of Giass, pp 1-7 Books for Industry, Meeting of the Society for Biomaterials (Charles- Magazines for Industry, Inc , New York. 1979 ton, SC. May 20-23, 1990),Vol Xlll Society for "L L Hench and D E Clark, "Physical Chem- Biomaterials, Algonquin. IL, 1990 istry of Glass Surfaces," J Non-Cryst Soiids, 28, "M M Walker, 'An Investigation into the Bond- 83 j1978) ing Mechanisms of Bioglass"; M S Thesis Univer- 62R W Douglas and T M El-Shamy, "Reactions sity of Florida. Gainesville, FL, 1977 of Glasses with Aqueous Solutions," d Am Ceram "' M Jarcho, calcium Phosphate Ceramics as Soc, 50 [l] 1-8 (1967) Hard Tissue Prosthetics." Ciin. Orthop Reiaf. "C M Jantzen, "Prediction of Glass Durability Res,, 157, 259 (1981) as a Function of Environmental Conditions." Mater "D F Williams. "The Biocompatibility and Clini- Res. SOC, 125, 143-59 (1988) cal Uses of Calcium Phosphate Ceramics". 1'4CM Jantzen, M J Polodinec and G G pp 43-66 in Biocompatibility of Tissue Anaiogs, Wicks, "Thermodynamic Approach to Prediction Vol /I Edited by D F Williams CRC Press, Boca of the Stability of Proposed Radwaste Glasses", Raton, FL, 1985 pp 491-95 in Advances in Ceramics. Vol 8,Nu- "'J E Davies, "The Use of Cell and Tissue Cul- dear Waste Management Edited by G G Wicks ture to Investigate Bone Cell Reactions to Bioac- and W A Ross American Ceramic Society, tive Materials", p 195 in Handbook of Bioactive Columbus, OH, 1984 Ceramics, Vol I, Bioactive Glasses and Glass- "B C Bunker, D R Tallant, T J. Headley, G.L Ceramics Edited by T Yamamuro. L L Hench Turner. and R J Kukpatrik, "The Structure of and J Wilson CRC Press, Boca Raton, FL 1990 Leached Sodium Borosilicate Glass," Phys Chem "J E Davies, ed Proceedings of the Bone- Giasses, 29 [3] 106-20 (1988) Biomaterials Interface Workshop (Toronto, Ontario, "A E Clark, C Y Kim. J K West, J Wilson, Canada, December 3-4. 1990) University of and L.L Hench, "Reactions of Fluoride- and Toronto Press, Toronto, Ontario, Canada, in press Nonfluoride-Containing Bioactive Glasses" p 73 "R Reck. S Storkel, and A Meyer, "Bioactive in Handbook of Bioactive Ceramics. Vol I, Bioac- Glass-Ceramics in Middle Ear Surgery. An 8-Year tive Glasses and Glass-Ceramics Edited by T Review' p 100 in Bioceramics. Materials Charac- Yamarnuro, L L. Hench, and J Wilson CRC teristics Versus in Vivo Behavior, Vol. 523 Edited Press, Boca Raton, FL, 1990 by P Ducheyne and J Lemons Annals of New "M Ogino, F Ohuchi and L L Hench. "Com- York Academy of Sciences, New York 1988 positional Dependence of the Formation of Cal- "G E Merwin. 'Review of Bioactive Materials

cium Phosphate Films on Bioglass,' J Biomed. tor 0 to I ogi c and Max i I I of ac i a 1 A pp I i c a t i on s " , Mater Res., 14, 55-64 (1980) pp 323-28 in Handbook of Bioactive Ceramics, "A.E Clark, C G Pantano. and L L Hench. Vol I, Bioceramic Glasses and Glass Ceramics 'Auger Spectroscopic Analysis of Bioglass Corro- Edited by T Yamamuro. L L Hench, and J sion Films, ' d. Am Ceram SOC, 59 [l-21 37-39 Wilson CRC Press, Boca Raton, FL, 1990 (1976) "E Douek. "Otologic Applications of Bioglass" 69T Kokubo, "Surface Chemistry of Bioactive Implants": in Proceedings of iVth lntemationai Glass-Ceramics,"J Non-Cryst Solids, 120, 138-51 Symposium on Bioceramics in Medicine (London, (1990) U.K Sept 10-11 1991) Edited by W Bonfield '"0H Andersson, "The Bioactivity of Silicate Butterworth-Helncmann. Ltd , London, U K , 1991, Glass". Ph D Dissertation Department of Chemi- in press July 1991 Bioceramics: From Concept to Clinic 1509

*8T Yamamuro, "Reconstruction of the Iliac Polyethylene as an Analogous Materials for Bone Crest with Bioactive Glass-Ceramic Prostheses"; Replacement", pp. 173-77 in Bioceramics Ma- pp 335-42 in Handbook of Bioactive Ceramics, terials Characteristics vs in Vivo Behavioi. Vol 523 Vol I, Bioactive Glasses and Glass Ceramics Ed- Edited by P Ducheyne and J.E Lemons Annals ited by T Yamamuro. L L. Hench, and J Wilson of New York Academy of Sciences. New York. CRC Press, Boca Raton, FL, 1990 1988 88 T Yamamuro. "Replacement of the Spine with '"'J C Behiri and W Bonfield. "Fracture Bioactive Glass-Ceramic Prosthesis", pp. 343-52 Mechanics of Bone The Effects of Density, Speci- in Handbook of Bioactive Ceramics, Vol I, Bioac- men Thickness, and Crack Velocity on Longitudi- tive Glasses and Glass-Ceramics Edited by T nal Fracture," J Biomech., 17, 25 (1984) Yarnamuro, L L. Hench, and J Wilson CRC 104W Bonfield, C Doyle, and K E Tanner, "In Press, Boca Raton, FL. 1990 Vivo Evaluation of Hydroxyapatite-Reinforced "H R Stanley, M.B Hall, A E Clark, G E Polyethylene Composites", p 153 in Biologicai and Turner, F Colaizzi, L.L Hench. and C King. Biomechanical Performance of Materiais. Edited by 'Alveolar Ridge Maintenance Using Endosseously P Christel, A Meunier, and A J C. Lee Elsevier, Placed Bioglass" (45% Cones)", to be published New York, 1986. in 1992 '@'A.F Tencer, R. L Woodard, J. Swenson, and "J Walliker. H Carson, E.E Douek, A Fourcin, K L. Brown, "Mechanical and Bone lngrowth Prop- and S Rosen. p 265 in Proceedings of Cochiear erties of a Polymer-Coated, Porous, Synthetic. lmpiant Symposium Edited by P. Banfai Duren, Coralline Hydroxyapatite Bone-Graft Materials", FRG, Sept 7-12, 1987 p 153 in Bioceramics Materiais Characteristics vs "M Jarcho, J I Kay, R.H Gumrnaer. and H P In Vivo Behavior. Vol 523. Edited by P Ducheyne

Drobeck. ' Tissue. Cellular, and Subcellular Events and J E Lemons Annals of New York Academy of at a Bone-Ceramic Hydroxylapatite Interface," Sciences, New York, 1988 J Bioeng, 1, 79 (1977) i06 T Kasuga, K Nakajima, T Uno, and M 91 M Ogiso, H Kaneda. J Arasaki, I. Tabata, Yoshida, "Bioactive Glass-Ceramic Composite and T Hidaka. "Epithelial Attachment and Bone Toughened by Tetragonal Zirconia"; pp 137-42 in Tissue Formation on the Surface of Hydroxylapa- Handbook of Bioactive Ceramics, Vol I, Bioce- tite Ceramics". p. 4 1 5 in Proceeding of 1st Worid ramic Glasses and Glass Composites Edited by Biomateriais Congress (Baden, Austria, 1980) T Yamamuro, L.L Hench, and J Wilson CRC European Society for Biomaterials, 1980 Press, Boca Raton, FL, 1990. 94 M Ogiso, H Kaneda, J Arasaki, K Ishida, M lo7P Ducheyne and L L. Hench, "The Process- Shiota, T Mitsuwa, T Tabata, Y Yamazaki, and T ing of Static Mechanical Properties and Metal Hidaka, "Hydroxyapatite Ceramic Implants under Fibre-Reinforced Bioglass," J. Mater Sci, 17, 595- Occlusal Function', p 54 in Transactions of the 606 (1982). '08E 4th Annual Meeting of the Society for Biomaterials Schepers, P Ducheyne, and M. DeClerk, (Rensellaer, NY, May 28-31, 1981). Society for Bio- "Interfacial Analysis of Fiber-Reinforced Glass materials. Algonquin, IL. 1981. Dental Root Implants," J. Biomed Mater Res., 23, 95H Aoki, Y Shin, and M. Akao, 'Sintered Hy- 735-52 (1989). droxyapatite for a Percutaneous Device", p 1 in '09J C. Bokros, U.S. Pat No 3526005, 1971 110 Advances in Biomateriais, Vol 6. Edited by P. J C. Bokros, "Carbon Biomedical Devices," Ducheyne, P Christel, A Meunier, and A.T.C. Lee Carbon, 15, 355-71 (1977). Elsevier, New York, 1986 "'A Haubold, H.S Shim, and J C Bokros, 96 N Yoshiyama, Y Shin, and H Aoki, "Hydroxyl- "Carbon in Medical Devices", pp 3-42 in Biocom- apatite Percutaneous Access Device in Peritoneal patibiiity of Clinicai lmpiant Materials, Vol /I CRC Dialysis", p 377 in Handbook of Bioactive Ceram- Press, Boca Raton, FL, 1981. ics, Vol II, Calcium Phosphate and Hydroxylapa- "*J D Bokros, "Carbon Biomedical Devices." tite Ceramics Edited by T Yamamuro, L L Carbon, 18, 355-71 (1977) Hench. and J Wilson. CRC Press, Boca Raton, '13A D Haubold, R A. Yapp. and J D. Bokros, FL, 1990 "Carbons", pp 95-101 in Concise Encyclopedia of 97T Tsuli H Aoki, Y Shin, and T Togawa, "Im- Medical and Dental Materiais Edited by D plantation in the Human Forearm of a Percuta- Williams Pergainon Press, New York. 1990 neous Device Made of Sintered Hydroxylapatite", '14S.D Cook, "In Vivo Evaluation of Hydroxyl- pp 383-76 in Handbook of Bioactive Ceramics, apatite Coatings for Orthopedic and Dental Appli- Vol II, Calcium Phosphate and Hydroxylapatite cations", in Proceedings of 3rd internationai Sym- Ceramics Edited by T. Yamamuro, L L Hench, posium on Bioceramics in Medicine (Rose Hulman and J Wilson. CRC Press, Boca Raton FL, 1990 Institute of Technology, Terre Haute, IN, Nov 18- "J Wilson, ' Preface', in Handbook of Bioactive 20, 1990) Edited by S F Hulbert Moore-Langen, Ceramics, Vols I and II. Edited by T Yamamuro. Terre Haute, IN, 1991, in press L L Hench, and J Wilson CRC Press, Boca ''% Boos, M Thirwell, R Blanchard, C Cryips, Raton, FL, 1990 M Herba. L Gonzales, L. Rosenthall, J Skelton, "C Doyle, "Bioactive Composites in Orthope- and G Borleau, presenied at 25th Annual Meeting dics", pp 195-208 in Handbook of Bioactive of the American Society Clinical Oncology, San Ceramics, Vol. II, Calcium Phosphate and Hydroxyl- Francisco, CA, May 21-23, 1989 apatite Ceramics Edited by T Yamamuro ''6D Day; personal communication, 1991 L L Hench, and J Wilson CRC Press, Boca '"K Ohara, M. Ikenaga, T Nakamura, and T Raton, FL, 1990 Yamamuro, "Heat-Generating Bioactive Ceram- 100 P Ducheyne and J F McGucken, Jr, ics," J, Biomed Mater,, 25, 357-65 (1991) "Composite Bioactive Ceramic-Metal Materials"; 118 H A Benghurri and P.K Bajpai, "Sustained Release of Steroid Hormones from Polylactic Acid pp 185-86 in Handbook of Bioactive Ceramics, Vol (I. Calcium Phosphate and Hydroxylapatite or Polycaprolactone-Impregnated Ceramics", Ceramics Edited by T Yamamuro, L.L. Hench, pp 93-110 in Handbook of Bioactive Ceramics. and J Wilson CRC Press, Boca Raton, FL, 1990 Vol 11, Calcium Phosphate and Hydroxylapatite "'U Soltez, "Ceramics in Composites. Review Ceramics Edited by T. Yarnamuro, L.L. Hench, and Current Status": pp 137-56 in Eioceramics. and J Wilson CRC Press, Boca Raton, FL. 1990 Materials Characteristics vs in Vivo Behavior, '"W R Lacefield and L L Hench. "The Bond- Vol. 523 Edited by P. Ducheyne and J. E. Lemons ing of Bioglass" to a Cobalt-Chromium Surgical Annals of New York Academy of Sciences, New Implant Alloy." Biomaterials, 7 [2] 104-108 1986 York, 1988 1211 F G Evans and A King, Biomechanical Stud- '@'W Bonfield, "Hydroxyapatite-Reinforced ies of the Muscuioskeltai System, pp 49-53 1510 Journal of the American Ceramcc Society - Hench Vol. 74, No. 7

Charles C Thomas Springfield IL 1961 Iz2R Van Audekercke and M Martens Me- "'W Bonfield Elasticity and Viscoelasticity of chanical Properties of Cancellous Bone Cortical Bone pp 43-60 in Natural and Living pp 89-98 in Natufai and Living Biomaterials Biomaterials Edited by G W Hastings and P Edited by G W Hastings and P Ducheyne Ducheyne CRC Press Boca Raton FL 1984 CRC Press Boca Raton FL 1984 0

he now serves as Graduate Research Professor of Materials Science and En- gineering, Director of the Bioglasso Re- search Center, and Co-Director of the Advanced Materials Research Center. Professor Hench discovered Bioglass@, the first synthetic material to bond with living tissues, in 1969. During 1978 to 1985, he and his students conducted a series of basic science studies on glases to immobilize high-level radioactive wastes, including the first deep ge- ological burial in Sweden. In 1980, Professor Hench initiated study of low- temperature sol-gel processing of glasses, ceramics, and composites, which led to new classes of silica materials for optics termed GelsiP Professor Hench's studies have re- sulted in more than 300 scientific pub- lications, 19 books, 20 U.S patents, and 20 foreign patents. He is a mem- ber of the American Ceramic Society, Larry L. Hench graduated from The Academy of Ceramics, and Society of Ohio State University in 1961 and 1964 Biomaterials; he has served in a vari- with B.S and Ph.D. degrees in ceramic ety of offices and has received awards engineering. He joined the faculty of the from these and other academic and University of Florida in 1964 where scientific institutions.