Bio Materials

CHAPTER

Biomaterials

AB in other surgical specialties, plastic surgery has beenrevolutionized by implantable devices. These advances have

been largely based on biomaterials science and engineering.Materials have been used for the fabrication of permanentimplants to replace diseased, injured, or congenitally absenttissuesand to restore tissue function.

In the future, however, materials will playa different role inplasticsurgery. Materials will also be used for the fabrication of

temporary, absorbable implants to facilitate the regenerationof tissue. This new role for materials is associated with novel

therapeutic approaches associated with tissue engineering.

HISTORYAND OVERVIEW

The changing role for materials in plastic surgery can be con-sidered through the various eras of the discipline. The 1930sthrough the 1970s might be considered the "Age of Devices."During these years a wide variety of implants were developedfor treating myriad problems. These devices revolutionizedplastic surgery and were fabricated from relatively few mate-rials: stainless steel, cobalt-chromium alloy, polymethyl-methacrylate (PMMA), and silicone.

Selection criteria for materials in implants were (1) strength,(2)biocompatibility, and (3) degradation resistance. The lattertwo specifications were often combined and referred to as

"inertness." Materials proposed initially for implants in plasticsurgery and for other surgical applications were generallyshown to display this property in nonmedical uses, especiallymetallic materials. Advantages and disadvantages of materialsforspecific applications often were determined only after theirintroduction into the clinic. Materials commended for a partic-ularclinical application because of one favorable property werefound to fail because of a deficiency in another property.

The 1980s and 1990s might be considered the "Age ofBiomaterials." New materials were introduced for the refabri-

cation of devices already in clinical use to improve their perfor-mance. Also, the development of other materials led to newtypesof implants.

Titanium alloy was introduced as an alternative to cobalt-

chromium alloy for fabrication of components with lowerstiffness to reduce bone loss associated with stress shielding.

Myron Spector

Synthetic calcium phosphate materials were introduced as bonegraft substitute materials. Research was also initiated on carbon

fiber-reinforced polymer composite materials to replace metalsin devices with even lower stiffness. Absorbable polymers weredeveloped to produce biodegradable devices for fracture fixa-tion and other applications.

Also during these decades, however, the limitations of

permanent biomaterials and devices became apparent, largelyas a result of clinical experience and follow-up. Device failurefrom specific deficiencies of materials prompted the search fornew substances that might improve the performance andextend the longevity of these devices.

More recently, a new class of the absorbable materials has

been developed to augment or replace graft materials fordefects in bone and certain soft tissues. These materials include

different forms of absorbable polymers used in devices forfracture fixation, as well as calcium phosphate substances andcollagen materials. In many respects these absorbable sub-stances and the experience with transplants in many surgicalspecialties have served as the foundation for the imminent

"Age of Tissue Engineering." Currently, porous absorbablesynthetic and natural polymers are being investigated asscaffolds to be seeded with cells in vitro for the production oftissue in the laboratory or for the preparation of implants tofacilitate tissue regeneration in vivo.

We also seem to be on the threshold of the "Age of GeneTherapy." Although experiments are investigating the efficacyof injecting gene fragments alone, it is likely that absorbable

materials may be effectively employed as delivery systems forthese genes. Also, the ex vivo transfection of cells with selected

genes is certain to combine gene therapy with tissue engineer-ing. Laboratory findings have been promising, but much workremains before the benefit/risk ratio of gene therapy for specificindications can be assessed.

Selection and development of a material for a specificindication are based on the following (Figure 19-1):

1. The device's function

2. The device's effect on the body3. The body's effect on the implantIssues related to the attachment of tissue to the implant

(i.e., incorporation of the device into the body) also areimportant.

239

240 PART I PRINCIPLES AND TECHNIQUES

Implant Selection or Design

Function of device

MechanicalChemicalElectricalThermalOptical

Effects ofdevice on

bodySurface Bulk

Bioadhesion(i.e., tissuebonding)?(Surface)

Yes No

Mechanical

Interpenetrating(tension, shear, compression)

Chain entanglementECM in interconnecting

pores (porosity)InterCligitating(shear, compression)

ECM in surface

irregularities (rugosity)

Nonreactive("inert")

HydrophobicHydrogel

Chemical

Proteinadsorption/binding

Mineral deposition

Effects of body on implantResistance to degradation?

Yes(permanent);No (absorbable)Mechanical

Fracture (strength)Wear (abrasive,fatigue)

ChemicalWear (adhesive)CorrosionOxidationHydrolysisEnzymolysisDissolution/precipitation

Figure 19-1. Issues related to selection or design of materials for fabrication of implantabledevices. ECM,Extracellular matrix.

This chapter provides a framework for understanding thecomposition and properties of the wide variety of materialsused for implantable devices in plastic surgety as they impactthe biologic response to these implants. Additional back-ground on biomaterials science and engineering can be foundin a recent text.89

COMPOSITION AND PROPERTIES

The term biomaterials generally refers to synthetic and treatednatural materials employed in the fabrication of implantabledevices that are used to replace or augment tissue or organfunction. An understanding of the physical properties of thesematerials is important for the judicious implementation ofimplants and provides for a realistic expectation of clinicalperformance.

The physical properties of materials result from theirchemical makeup. Therefore an understanding of the molecu-lar structure and chemical bonding ofbiomaterials can providethe basis for understanding the physical properties.63

Metals

In metals, closely packed arrays of positively charged atoms areheld together in a loosely associated "cloud" of free electrons.The essential features of the metallic bond are that (1) it is

nondirectional and (2) the electrons are freely mobile. The

metals most often used for the fabrication of implantabledevices are stainless steel, cobalt-chromium alloys, and tita-nium and titanium alloy. The specific members of thesefamilies used as biomaterials are usually identified by a desig-nation provided by the American Society for Testing andMaterials (ASTM).

The stainless steels, as with all steels, are iron-based alloys.Chromium is added to improve the corrosion resistancethrough the formation of a chromium oxide surface layer; atleast 17% chromium is required for the term stainless to beused. Carbon and nickel are employed as alloying elements toincrease strength. The most common type of stainless steelused for implants is 316L (American Iron and Steel Institutedesignation; ASTM F-138), containing 17% to 19% chro-mium, 13% to 15.5% nickel, and less than 0.03% carbon.

Surgical cobalt-chromium alloy is a cobalt-based system withchromium added for increased corrosion resistance. Its compo-sition contains 27% to 30% chromium and 5% to 7% molyb-denum. Tungsten is added to the wrought alloy to enhanceductility.

Titanium and its alloy, with 6% aluminum and 4%vanadium (Ti-6Al-4V), are used for their excellent corrosion

resistance and their modulus of elasticity, which is approxi-mately one-half that of stainless steel and cobalt-chromiumalloys. This lower modulus results in devices with lowerstiffness, which may be advantageous in certain applications,such as implants in bone because they will result in less stress

CHAPTER 19 Biomaterials 241

shielding of bone. The alloy of titanium has much better mate-rial properties than the pure titanium. Problems with titaniumare its severe notch sensitivity and poor wear resistance.

The mechanical properties of any material are linked to itsultrastructure.Any processing treatment that alters the phasestructure, grain size, or grain orientation of a metal will affectthe material properties. Mechanically deforming a metal be-yond its yield stress will produce, in addition to a shape change,a material with increased yield and ultimate stresses and ahigher endurance limit. Because this process, called cold work-ing, results in permanent deformation, the resultant material islessductile than the initial material.

Casting is the pouring of molten metal into a mold toproduce a specific shape on cooling. Voids and other Rawsare amajor problem with casting. Forging is a process by which ablank of metal is heated and pressed into a die by the applica-tion of a large, single force. This produces a part with fewerdefects than a casting. Annealing is essentially heating for pro-longed periods to allow stress relaxation and grain reorganiza-tion and growth. Hot isostaticpressing is a newer technologyinvolving the consolidation of metal powder into a fine-grainedmaterial under high temperature and pressure. These and otherprocessing methods have been developed to produce parts ofcomplex shape with the highest degree of mechanical integrity.

Metallic materials have certain properties that make themideal for load-bearing applications. They can maintain veryhigh strength under the aggressive aqueous environment in thebody. The biocompatibility of metallic materials is related totheir corrosion resistance.

STAINLESS STEELS. Many alloys and grades of alloys areidentified as commercial stainless steels. Only the iron-chromium-nickel alloys are used as biomaterials. Unalloyediron, carbon steels, and alloyed carbon steels cannot be usedbecausethey corrode in oxygenated saline solutions.

Despite their very good corrosion resistance, stainless steelsare subject to several other corrosion processes, includingcrevice, pitting, intergranular, and stress corrosion. Theseprocessescan profoundly degrade the mechanical strength ofthe alloy and can lead to the release of metallic ions into thesurrounding tissue, with undesirable biologic consequences(seelater discussion).

Alloying with chromium generates a protective, self-regenerating oxide film that resists perforation, has a highdegree of electrical resistivity, and thus provides a majorprotection against corrosion. Formation of the chromium.oxide "passivation" layer is facilitated by immersion of thealloyin a strong nitric acid soliltion. The nickel imparts morecorrosion resistance and outstanding fabricability. The molyb-denum addition provides resistance to pitting corrosion. Otheralloyingelements facilitate manufacturing processes.

The presence of carbon in stainless steel is undesirable.Under certain conditions the carbon segregates from the majorelements of the alloy, taking with it a substantial amount ofchromium in forming chromium carbide precipitates. Localdepletion of chromium deprives those zones of corrosionresistance,and since the carbides form most frequently at the

alloy crystal interfaces, the resultant corrosion occurs selec-

tively in the intercrystalline paths. Certain grades of stainlesssteel (e.g., F-138) have a maximum of only 0.03% carbon toreduce the formation of these carbides.

Another feature of the stainless steels relates to the presenceof inclusions, which can serve as Raws or cracks that can be

propagated by cyclic loading-the process offatigue failure. Toincrease resistance to fatigue failure, certain grades of steel areavailable with smaller and more widely spaced inclusions.

COBALT-CHROMIUM ALLOYS. As with stainless steels, the

chromium content of this alloy generates a highly resistive,passive film that contributes substantially to corrosion resis-tance. The cobalt-chromium-molybdenum (Co-Cr-Mo) alloy(F-75) has superior corrosion resistance to the F-138 stainlesssteel, particularly in crevice corrosion, and has ari extensivehistory ofbiocompatibility in human implantation.

Co-Cr-Mo devices are currently produced by hot isostaticpressing, which results in parts with more favorable strengthcharacteristics than results from casting processes. In the hotisostatic pressing process the liquid alloy is atomized topowder. Loose powder is consolidated to a void-free solid,resulting in a part reasonably close in shape to the ultimatedevice. The preformed part is finished by machining andpolishing. The resulting structure has a very fine, alloy crystalsize and very fine, uniform dispersion of carbide particles. Theavailability of the high-strength varieties and their superiorcorrosion resistance and biocompatibility make them excellentchoices for high-stress applications.

Cobalt-chromium-titanium-nickel {Co-Cr- Ti-NOalloy(ASTM F-90) is very different from the F-75 alloy, with which

it is often confused. This alloy can be hot forged and colddrawn and is not used in the cast form. In clinical practice it isused to make wire and internal fixation devices (e.g., plates,intramedullary rods, screws).

UNALLOYED TITANIUM AND TITANIUM ALLOY. Tita-

nium and its alloys are of particular interest for biomedicalapplications because of their outstanding biocompatibility. Ingeneral, their corrosion resistance significantly exceeds that ofthe stainless steels and the cobalt-chromium alloys. In salinesolutions at near-neutral pH the corrosion rate is extremelysmall, with no evidence of pitting, intergranular, or crevicecorrosion. Data from in vivo animal experimental models andfrom human sources indicate superior biocompatibility.

So-called unalloyed titanium actually is alloyed by thelevel of oxygen dissolved into the metal. In large amounts,oxygen embrittles titanium and its alloys, but in small,regulated amounts it helps control the yield strength ofthe materials. ASTM F-67 is a specification for oxygenproviding 345 megapascals (MPa) of yield strength in gradeIII and 485 MPa of minimum yield strength in grade IV.Other elements in unalloyed titanium include nitrogen,0.07% (maximum); carbon, 0.15% (maximum); hydrogen,0.015% (maximum); and iron, 0.35% (maximum). Anyexcess of these elements may degrade the performance of thebasic material.


Unalloyed titanium is used less frequendy than the alloy forimplants but is available in various configurations, such asplain wire for manufacturing purposes. In addition, it is usedto produce porous coatings for certain designs of total jointreplacement prostheses.

ASTM F-I36 specifies a titanium alloy with a content of5.5% to 6.5% aluminum, 3.5% to 4.5% vanadium, 0.25%

iron (maximum), 0.05% nitrogen (maximum), 0.08% carbon(maximum), 0.0125% hydrogen (maximum), and 0.1% other(maximum 0.4% total). Oeveloped by the aircraft industry asone of several high-strength titanium alloys, this particular for-m ulation has a yield strength reaching 1110 MPa. The ASTMF-I36 specification limits the oxygen to an especially low levelof 0.13% maximum. This is also known in the industry as theextra low interstitial (ELI) grade. Limiting the level of oxygenimproves the mechanical properties of the material, particularlyincreasing its fatigue life.

One interesting feature of titanium and its alloys is the lowmodulus of elasticity of 100 gigapascals (GPa), as comparedwith 200 GPa for the cobalt-chromium alloys. This featureleads to their use in plates for internal fixation of fractUres.Some have found that the lower stiffness of these plates maydecrease the severity of bone stress shielding, which results inosteopenia under these devices.

One of the weaknesses of titanium is poor wear resistance.This problem ~'parendy relates to the mechanical stabiliryof the passive film covering the alloy's surface. On a carefullypolished surface, the film is highly passive but mechani-cally weak.

Permanent and Absorbable Syntheticand Natural PolymersPolymers consist of long chains of covalendy bonded mol-ecules characterized by the repeated appearance of a mono-meric molecular unit. They can be produced de novo by thepolymerization of synthetic monomers or prepared fromnatural polymers isolated from tissues. Most synthetic andnatural polymers have a carbon backbone. Bonding amongpolymer chains results from the much weaker secondary forcesof hydrogen bonds or van der Waal' s forces. Covalent bondingamong chains, referred to as cross-linking, can be produced incertain polymer systems.

Physical entanglements of the long polymer chains, thedegree of crystallinity, and chemical cross-linking amongchains play important roles in determining polymer properties.The molecular bonding of the backbone of the polymer can bedesigned to undergo hydrolysis or enzymatic breakdown, thusallowing for the synthesis of absorbable, as well as permanent,devices.

Polymeric materials are generally employed for the fabri-cation of implants for soft tissue applications that require agreater degree of compliance than can be achieved with metals.They have also been shown to be of value as implants in bonefor indications that would also benefit from their lower modu-

lus of elasticity, as well as the ability of some to be polymerizedin vivo and adapt to defects of complex shape. For someindications the radiolucency of polymeric materials may be an

important benefit. Because of the limited strength and wearresistance of polymers, the load-bearing requirements of theapplications must be considered.

POLYMETHYLMETHACRYLATE. PMMA is used in a self-

curing form as a filling materials for defects in bone and as agrouting agent for joint replacement prostheses.35 It can beshaped in vivo while in a dough stage before complete polymer-ization, and thus a custom implant is created for each use. Itspurpose is more even redistribution of stress on the surround-ing bone.

Often referred to as "bone cement," when PMMA is em-

ployed for joint arthroplasty, it acts as a grout to support theprosthesis rather than a glue; it has minimal adhesive proper-ties. The time-dependent properties of PMMA during curingrequire an understanding ofits handling characteristics. Imme-diately after mixing, the low viscosity permits interdigitationwith cancellous bone. Viscosity rises quickly once the chemicalsetting reaction begins, requiring that the prosthesis be accu-rately positioned and stationary to achieve maximum fixation.

The chemical toxicity of the methylmethacrylate monomerand the heat generated during polymerization must be consid-ered when using PMMA. Its britde nature after curing and lowfatigue strength make PMMA vulnerable to fractUre under highmechanical loading. Also, wear debris is produced when other,harder materials rub ~ainst PMMA.

Although the adjunctive use of PMMA to stabilize patho-logic fractures of diaphyseal bones has become routine, itsimplantation about the cervical spine has been limited becauseof the exothermic reaction during the curing process. Bone-cement interfacial temperatUres of 600 to 650 C (1400 to1490 F) during total knee arthroplasty and of 800 to 900 C(1760 to 1940 F) during total hip arthroplasty have beenreported. These temperatures are of sufficient magnitUde tocause degradation of protein. The adverse effects of such tem-peratures on the cerebrospinal fluid and spinal cord can beavoided by the following:

1. Precooling the soft tissues and bone2. Placing the cement on the dorsal aspect of the intact

lamina

3. Applying precooled Gelfoam over exposed duraIn this way, acrylic cement and wire fixation of pathologic

fractures of the cervical spine can be performed safely to allowearly mobilization of patients with limited life expectancies.

SILICONE. Silicones are polymers with a backbone comprisedof alternating silicon and oxygen atoms and organic side groupsbonded to the silicon through covalent bonding with the car-bon atom.61 One form of silicone typically used for the fabrica-tion of implants is polydimethylsiloxane (POMS). In POMS,methyl (CH3) side groups are covalendy bonded to the siliconatom. POMS can be used in the following three forms:

1. A fluid comprising linear polymers of varying molecularweight (i.e., chain length)

2. A cross-linked network referred to as a gel3. A solid elastomer comprising a highly cross-linked gel

filled with small particles of silica.


Silica is another silicon-containing molecule, silicon diox-ide. PDMS elastomers contain a nonctystalline silica particle 7to 22 nm in diameter that has been surface-treated to facilitate

chemical bonding of the particle to the PDMS gel. Addition ofthe silica particle to a highly cross-linked PDMS gel is done tomodify the mechanical properties of the elastomer. In certaincases, such as the PDMS elastomeric shell for silicone-filled

breast implants, the elastomer may contain side groups ofphenyl (C6Hs) or trifluoropropyl (CF3CH2CH2) instead ofthe methyl groups to better contain the PDMS fluid.

In evaluating the performance of silicone implants, the roleof each form of PDMS in the device must be considered.

Attributing specific biologic responses to individual compo-nents of a silicone device is complicated by the implant's manymolecular forms of silicone.

POLYETHYLENE. Ultrahigh-molecular-weight polyethylene(UHMWPE) has a vety low frictional coefficient against metaland ceramics and is therefore used as a bearing surface for jointreplacement prostheses.13,62 Moreover, the wear resistance ofUHMWPE is lower than that of other polymers investigatedfor this application. Low strength and creep, however, presentpotential problems.

The term polyethylene refers to plastics formed from poly-merization of ethylene gas. The possibilities for structural vari-ation in molecules formed by this simple repeating unit fordifferent factors (e.g., molecular weight, ctystallinity, branch-ing, cross-linking) are so numerous and dramatic with sucha wide range of attainable properties that polyethylene trulyrefersto a subclass of materials. The earliest type of polyethyl-ene was made by reacting ethylene at high (20,000 to 30,000pounds per square inch) pressure and temperatures of 2000 to4000 C with oxygen as a catalyst. Such material is referred to asconventional or low-density, polyethylene.

Much polyethylene is produced now by newer, low-pressure techniques using aluminum-titanium (Ziegler) cata-lysts.This is called linear polyethylene due to the linearity of itsmolecules, in contrast to the branched molecules produced byhigh-pressure processes. The linear polymers can be used tomake high-density polyethylene by means of the higher degreeof ctystallinity attained with the regularly shaped molecules.Typically, molecular weight is not significantly differentbetween the low-density and high-density varieties (e.g.,100,000 to 500,000).

If the low-pressure process is used to make extremely longmolecules (i.e., UHMWPE), however, the result is remarkablydifferent. This material, with a molecular weight of 1 to 10million, is less ctystalline and less dense than high-densitypolyethylene and has exceptional mechanical properties. It isextremely tough and is remarkably wear resistant; a 0.357-magnum bullet fired from 25 feet bounces back from al-inch-thick slab of UHMWPE. The material is used in vety

demanding applications (e.g., ore chutes in mining equip-ment) and is the most successful polymer used in total jointreplacements. It far outperforms the various actylics, fluorocar-bons, polyacetals, polyamides, and polyesters tried for suchpurposes.

ABSORBABLE POLYMERS

Synthetic. Absorbable polymers have been used in plasticsurgety for decades in the form of absorbable sutures.4 Morerecently this class of materials has been investigated for theapplication of resorbable devices, including fracture fixationimplants and scaffolds for tissue engineering. The principalissues associated with the implementation of absorbablepolymers as implants include the following:

1. Mechanical properties (e.g., strength)2. Degradation rate3. Biologic response to the degradation products.One of the classes of polymers used frequently for the

fabrication of absorbable implants is the alpha-hydroxy acids,including L-Iactic acid, glycolic acid, and dioxanone. Thesemolecules normally are used in their polymeric forms: polY-L-lactic acid (PLLA), polyglycolic acid (PGA), and polydiox-anone. Copolymers of lactic and glycolic acids are also fre-quently employed.

A particular class of polyester undergoes breakdown as aresult of the hydrolytic scission of the ester bond. The access ofwater to this bond in PGA is much greater, resulting in a morerapid degradation rate compared with that occurring withPLLA, which has a bulkier CH3 side group instead of the Hatom in PGA. The copolymer of polylactic and polyglycolicacid can be designed to have an intermediate degradation rate.Although most breakdowns of these polymers are caused byhydrolytic scission, nonspecific enzymatic action is alsoinvolved.

Factors that affect the rate of breakdown of these polymersinclude (1) the relative amount of monomers composing thecopolymers, (2) the degree of ctystallinity, and (3) the surfacearea. These polymers are normally broken down to naturalbody components excreted in the urine or exhaled. The processof degradation involves the gradual decrease in the averagemolecular weight of the polymer as hydrolysis proceeds. Atsome point the molecular weight decreases to the extent thatthe polymer becomes soluble in the aqueous environment, anda bolus release of the molecules occurs. Depending on the massof the implanted device, the concentration of the moleculesmay elicit an inflammatoty response. 108

NaturaL Myriad devices are fabricated from collagen, theprincipal structural protein of the body. 117The collagen mole-cule comprises three tightly coiled helical polypeptide chains.In vivo the collagen molecule, tropocollagen, is assembled toform fibrils, which in turn assume various orientations and

configurations to form the architecture of various tissues. Thewide array of properties of tissues that make up collagen, fromdermis to musculoskeletal tissues and including articular carti-lage, meniscus, and ligament, is caused by differences in thechemistty, density, and orientation of the fibrils formed fromthe collagen molecule.

Collagen is soluble in specific solutions in which the chainscan become disentangled to produce gelatin. Collagen can beisolated from tissue and purified through the use of severalagents: acids, alkalis, enzymes, and salt. Treatment in acidresults in the elimination of acidic proteins and glycosami-

.244 PART I PRINCIPLES AND TECHNIQUES

noglycans, which cause dissociation of the collagen fibrils. Asimilar effect can be achieved using alkaline extraction with theremoval of basic proteins. Proteolytic enzymes that cleave thetelopeptides, which serve as natural cross-linking agents forcollagen, allow for the dissolution of collagen molecules andaggregates in aqueous solutions. Salt extraction leads to theremoval of newly synthesized collagen molecules and cenainnoncollagenous molecules, thus facilitating the disaggregationof collagen fibrils.

Collagen's solubility in an acidic medium facilitates its ex-traction from tissues and reprocessing into biomaterials. Severalfactors are critical determinants of the propenies of reconsti-tuted collagen biomaterials. The degree of denaturation ordegradation of the collagen structures isolated from tissue willaffect the mechanical properties. These properties will also beaffected by the degree to which the material is subsequentlycross-linked.

An imponant biologic property related to collagen's molec-ular structure is the collagen-induced bloodplatelet aggregation.The quarternary structure of collagen resulting from the per-iodic aggregation of the collagen molecules has been well docu-mented. Methods for isolating and purifying collagen fibrilsresult in the preservation or destruction of this quarternarystructure and are used to produce either hemostatic or throm-boresistant biomaterials. Another factor relates to the removal

of soluble components that might serve as antigens. The immu-nogemclty can be reduced' to clinica.l1yinsignifiCant levels bychemically modifying the antigen molecules.

A wide variety of methods have been employed for thefabrication of collagen sutures, fleeces for hemostasis, andspongelike materials for scaffolds in tissue engineering.

Ceramics

Ceramics are typically three-dimensional arrays of positivelycharged metal ions and negatively charged nonmetal ions, oftenoxygen. The ionic bond localizes all the available electrons inthe formation of a bond. Network organization ranges fromhighly organized, crystalline, three-dimensional arrays to amor-phous, random arrangements in glassy materials.

Ceramics may be the most chemically inert implant materi-als currently in use. Their relatively low tensile strength, highmodulus, and brittleness, however, limit their applications.Current techniques allowing the formation of ceramic coatingon metallic substrates have revitalized interest in ceramics for

hard-tissue applications.

ALUMINUM OXIDE. Aluminum oxide has been found of

value for the aniculating components of total joint arthroplas-ties because ofits high wear resistance and its low coefficient offriction when prepared in congruent, polished geometrics. Thebrittle nature of alumina remains a detriment.

CALCIUM PHOSPHATES AND HYDROXYAPATITE.

Calcium-based ceramics, closely related to the naturallyoccur-ring hydroxyapatite in bone, have generated great interest inrecent years. Their ability to bond directly to bone and theirosteoconductive capability promise to enhance biologic fixa-tion of implant devices. Hydroxyapatite is only slightly resorb-

able and is used in both dense and porous forms as a perma-nent implant. Tricalcium phosphate is bioabsorbable tovarying degrees, depending on formulation and structure.Many calcium phosphate materials are currently under-going investigation as bone graft substitute materials (see laterdiscussion).

Composite MaterialsComposite materials are combinations of two or more materialsand usually involve more than one material class (e.g., metals,polymers, ceramics). They are used to achieve a combinationof mechanical propenies for specific applications. Compositetechnology, much of it developed for the aerospace industry, isbeginning to make its way into biomedical materials. Carbonfiber-reinforced polymers are being investigated as substitutesfor metals. The advantage is that devices with comparablestrength but with significantly lower stiffness can be produced.Moreover, these types of composite devices are radiolucent.

BIOLOGIC RESPONSE TO BIOMATERIALS

The biological processes composing the tissue response areaffected by the following implant-related factors99:

1. "Dead space" created by presence of the implant2. Soluble agents released by the i"!plant (e.:J1;.,metal ions,

polymer fragments)3. Insoluble paniculate material released from the implant

(e.g., wear debris)4. Chemical interactions of biologic molecules with the

implant's surface5. Alterations in the strain distribution in tissue caused by

(a) mismatch in the modulus of elasticity between theimplant and surrounding tissue and (b) movement ofthe implant relative to adjacent tissue as a result of theabsence of mechanical continuity

Study of the tissue response to implants requires methodol-ogy capable of measurements at the molecular, cellular, andtissue levels (Figure 19-2). Time is an impottant variable be-cause of the critical temporal relationship between the molecu-lar and cellular protagonists of the biologic reactions. Also,implant-related factors act on the biologic responses with dif-ferent time constants. The dynamic nature of implant-tissueinteractions requires that the final assessment of tissue compat-ibility be qualified by the time frame in which it has beenevaluated.

The tissue response to an implant is the cumulative physio-logic effect of the following 10°:

1. Modulation of the acute wound healing response to thesurgical trauma of implantation and presence of theimplant (Figure 19-3)

2. Subsequent chronic inflammatory reaction associatedwith presence of the device

3. Remodeling of surrounding tissue as it adapts to pres-ence of the implant

In addition, the healing and stress-induced adaptive remod-eling responses of different tissues to the same implant can varygreatly.


I flm

: !: : Bone

i Proteins i Mh

i ~cr : "'mp age : ~

Particles ic:;::~~ I Osteoblasts

c D<J ,/---;! ,%o D: 0° : ~

0,,68 ~' ~

: \: . ~

.::: i

@;

!.. ~

Ion,: . ; '"a,leoda,'Fibroblast ... i~ibrous. ... :lis sue (sea )..' r.+.+ :...

11

'

co agen:.;;;?'".

Biomaterial

Protein adsorption/conformational change/

desorption

nm

WearMetal corrosion

Polymer degradation

Cell and Tissue Response

Size scalemm

Figure 19-2.tissues.

Interactions among moieties released by implants, biologic molecules, cells, and

CVascular response

Clotting~cute . Phagocytosis .InRammahon Neovascularizati~ <?ranulahon

New collagen synth~~is ~ hssue

~ -------Tissueof labile Tissuesofor stablecells permanentcells

~ ~-Imprantmovement f.Framework Framework Scarring

intact destroyed Fibrousencapsulation~ ~ Ch .. ~R .

Regeneration Scarring romcIn ammahon(incorporationof Fibrousencapsulationimplantintotissue) ~

ChronicinRammation(synovium,granuloma)

Figure 19-3. Biologicprocesses initiated by surgicaltrauma of implantation that determine tissueresponse to implantable devices.

The biologic response elicited by an implant involves thehealingand remodeling characteristics of the four basic typesoftissue:connective tissue, muscle, epithelium, and nerve. Thecharacteristicsof the parenchymal cells in each type of tissuecanprovide a basis for understanding the tissue response to animplant. The following characteristics of an implant site aredeterminants of the biologic response:

1. Vascularity2. Nature of the parenchymal cell with respect to its capa-

bility for mitosis and migration, because these processesdetermine the regenerative capability of the tissue

3. Presence of regulatory cells, such as macrophages andhistiocytes

4. Effect of mechanical strain, associated with deformationof the extracellular matrix, on the behavior of the

parenchymal cell

Surgical wounds in avascular tissue (e.g., cornea, inner thirdof meniscus) will not heal because of the limited potential forthe proliferation and migration of surrounding parenchymalcells into the wound site. Gaps between an implant and sur-rounding avascular tissue will remain indefinitely. Implant sitesin vascular tissues where the parenchymal cell is incapable ofmitosis (e.g., nerve tissue) will heal by scar formation in the gapbetween the implant and surrounding tissue. Moreover, adja-cent cells that have died as a result of the implant surgery willbe replaced by fibroblasts and scar tissue.

Normal Local Tissue ResponseWOUND HEALING. The tissue that forms around implantsis the result of the influence of the device on (1) the wound

healing response initiated by the surgical trauma of implanta-tion (Figure 19-3) and (2) subsequent tissue remodeling.


Implantation of a medical device initiates a sequence of cellularand biochemical processes that lead to healing by secondaryintention, that is, healing by the formation of granulation tissuewithin a defect, as opposed to the healing of an incision, or,healing by primary intention.

The first phase of healing in vascularized tissues is inflamma-tion. This is followed by a reparative phase, the replacement ofthe dead or damaged cells by healthy cells. The pathway of thereparative process depends on the regenerative capability of thecells composing the injured tissue (i.e., the tissue or organ intowhich the implant has been placed).

Cells can be distinguished as labile, stable, or permanentbased on their capacity to regenerate. Labile cells continue toproliferate throughout life, replacing cells that are continuallybeing destroyed. Epithelial and blood cells are examples oflabile cells, as are cells of splenic, lymphoid, and hematopoietictissues.

Stable cells retain the capacity for proliferation, althoughthey do not normally replicate. These cells can undergo rapiddivision in response to a variety of stimuli and can reconstitutethe tissue of origin. Stable cells include the parenchymal cellsof all the glandular organs of the body (e.g., liver, kidney,pancreas), mesenchymal derivatives (e.g., fibroblasts), smoothmuscle cells, osteoblasts and chondrocytes, and vascularendothelial cells.

Permanent cells cannot reproduce themselves after birth.Examples are nerve cells.

Tissues comprised of labile and stable cells can regenerateafter surgical trauma (Figure 19-3). The injured tissue isreplaced by parenchymal cells of the same type, often leavingno residual trace of injury. Tissues comprised of permanentcells, however, are repaired by the production of fibrocollag-enous scar. Despite the capability of many tissues to undergoregeneration, destruction of the tissue stroma remaining afterinjury or constructed during the healing process will lead toformation of scar. The biologic response to materials thereforedepends on the influence of the material on the inflammatoryand reparative stages of wound healing.

The following types of questions need to be addressed whenassessing the "biocompatibility" of materials:

Does the material yield leachables or corrosion products thatinterfere with the resolution of inflammation initiated bythe surgical trauma?

Does the presence of the material interfere with the stromarequired for the regeneration of tissue at the implant site?

A number of systemic and local factors influence theinflammatory-reparative response. Systemic influences includeage, nutrition, hematologic derangements, metabolic derange-ments, hormones, and steroids. Although the prevailing "con-ventional wisdom" is that elderly persons heal more slowly thanyoung people, few control data and animal experiments sup-port this theory. Nutrition can have a profound effect on thehealing of wounds. Prolonged protein starvation can inhibitcollagen formation, and high-protein diets can enhance the rateof tensile strength gained during wound healing. Local influ-ences that can affect wound healing include infection, inade-quate blood supply, and presence of a foreign body.

FIBROUS TISSUE INTERFACE. The implant creates a deadspace in tissue that attracts macrophages to the implant-tissueinterface.96 These cells are attracted to the prosthesis as they areto any dead space (e.g., bursa, joint space), presumably becauseof certain microenvironmental conditions (e.g., low oxygen,high lactate). In this regard it is not clear why macrophages areabsent from the surface of osseointegrated implants (see follow-ing discussion).

Macrophages and the scar's fibroblasts form synovial tissue,which can be considered the chronic inflammatory response toimplants, unless the device is apposed by osseous tissue (i.e.,osseointegrated).101 This process is often termed fibrous encap-sulation.27,54The presence of regulatory cells such as macro-phages at the implant-tissue interface can profoundly influencethe host response to a device because these cells can releaseproinflammatory mediators if irritated by the movement of thedevice or substances released from the biomaterial.3 The in-

flammatory response of the synovial tissue around implants iscomparable to the inflammation that can occur in the syn-ovium lining any bursa (e.g., bursitis); thus the response to.

I h b d.

Ia b .. 101Imp ants as een terme Imp nt urslns.

RESPONSE TO IMPLANTS IN BONE. Wound healinggoverns the makeup of the tissue that forms around implants.Because it can regenerate, bone should be expected to apposeimplants in osseous tissue and form within the pore spaces ofporous coatings. Is this bone bonded in any way to theimplant? Bonding of a prosthesis to bone would enhance itsstability, limiting the relative motion between the implant andbone. In addition, bonding might provide a more favorabledistribution of stress to surrounding osseous tissue.

Bonding and Osseointegration. Bonding of bone to animplant can be achieved by mechanical or chemical means.Interdigitation of bone with PMMA bone cement or withirregularities in implant topography and bone ingrowth intoporous surfaces can yield interfaces capable of supporting shearand tensile as well as compressive forces. These types ofmechanical bonding have been extensively investigated and arereasonably well understood.

Chemical bonding of bone to materials may result frommolecular (e.g., protein) adsorption and bonding to surfaceswith subsequent bone cell attachment. This phenomenon hasundergone intensive investigation in recent years but is not yetas well understood as mechanical bonding.

The term osseointegration has been used to describe thepresence of bone on the surface of an implant with no histologicintervening nonosseous (e.g., fibrous) tissue. All implants inbone should become osseointegrated unless the bone regenera-tion process is inhibited.

Bone Ingrowth. The bone ingrowth into a porous-surfacecoating on an implant leads to an interlocking bond thatcan stabilize the device. In order for the porous materialto accommodate the cellular and extracellular elements of

bone, the average pore diameter should be greater thanabout 100 ~m.

CHAPTER19 Biomaterials 247

The bone ingrowth process proceeds in two stages. First,surgical trauma of implantation initially leads to the regenera-tion of bone throughout the pores of the coating. Second,mechanical stress-induced remodeling leads to resorption ofbone from certain regions of the implant and continuedformation and remodeling of bone in other regions.

Chemical Bonding of Bone to Biomaterial Previous

investigations have provided evidence of bone bonding tomany different types of calcium phosphate materials, calciumcarbonate substances, and calcium-containing "bioactive"glasses.Chemical bonding was evidenced by the high strengthof the implant-bone interface, which could not be explained bya mechanical interlocking bond alone. In addition, electronmicroscopy has shown that no identifiable border exists be-tween these calcium-containing implants and adjacent bone.

Many recent studies have investigated the bonding of boneto one particular calcium phosphate mineral, hydroxyapatite.This was chosen because of its relationship to the primarymineral constituent of bone; natural bone mineral is a calcium-

deficient carbonate apatite. Experiments have been performedon both hydroxyapatite-coated metallic implants and on partic-ulate and block forms of the mineral used as bone substitute

materials. Histology of specimens shows that a layer of newbone approximately 100 ~m in thickness covers most of thehydroxyapatite sutface within a few weeks of implantation andremains indefinitely. This layer of bone is attached to thesurrounding osseous tissue by trabecular bridges.

In studying the mechanism of bone bonding, researchershave found that within days of implantation, biologic apatitesprecipitate (from body fluid) onto the surface of the calcium-containing implants. These biologic apatites are comparable tothe carbonate apatite of bone mineral. Proteins probably adsorbto this biologic mineral layer, thereby facilitating bone cellattachment and the production of osteoid directly onto theimplant. This osteoid subsequently undergoes mineralization,as it does normally in osteogenesis, thus forming a continuumof mineral from the implant to the bone. The bone cell re-sponds to the biologic apatite layer that has formed on theimplant and not directly to the implant itself.

Recent studies have shown that this biologic apatite layerforms on many different calcium phosphate substances, ex-plaining why bone-bonding behavior has been reported formany different types of calcium phosphate materials. The clin-ical value of this phenomenon will depend on the following:

1. For coatings, how well these substances can be bondedto implants

2. For bone graft substitute materials, their strength,modulus of elasticity, and ability to be resorbed

The finding that bone can become chemically bonded tocertain biomaterials, however, is a significant advance in under-standing the implant-bone interface.

IMPLANT-INDUCED ALTERATIONS. The presence of theimplant can alter the stress distribution in the extracellular

matrix and thereby reduce or increase the strains experiencedby the constituent cells. Many studies have demonstrated

immobilization-induced atrophy of certain tissues resultingfrom the decrease in mechanical strains. Loss of bone mass

around stiff femoral stems and femoral condylar prostheses oftotal hip and knee replacement devices has been associatedwith the reduced strains resulting from stress shielding.Hyperplasia and hypertrophy of tissue have also occurred, in

which mechanical strains have increased due to the presence ofan implant.

CRITERIA FOR TISSUE RESPONSE. The in vivo assessment

of tissue compatibility of biomaterials requires that certaincriteria be implemented for determining the acceptability ofthe tissue response relative to the intended application of thematerial or device. The biomaterial or device should be

considered biocompatible only in this context. Every studyinvolving the in vivo assessment of tissue compatibility shouldprovide a working definition ofbiocompatibility.

Biomaterials and devices implanted into bone can becomeapposed by the regenerating osseous tissue and thus can beconsidered compatible with bone regeneration. Altered boneremodeling around the device caused by stress shielding, with anet loss of bone mass (i.e., osteopenia), could lead to theassessment that the material or device is not compatible withnormal bone remodeling. When the implant is surrounded byfibrous tissue, macrophages appearing on the material's surfaceare the expected response to the dead space produced by theimplant. The synovial tissue thus produced might be consid-ered an acceptable response relative to the material's chemicalcompatibility. Using the thickness of the scar capsule aroundimplants alone as a measure of biocompatibility is problematicbecause it can be influenced by tissue movement at the siterelative to the implant.

The cellular and molecular makeup of tissue and the inter-actions among these components are complex. Criteria forassessing certain features of the biocompatibility ofbiomaterialsand devices should focus on specific aspects of the biologicresponse. Importantly, materials yielding acceptable tissuecompatibility in one site of implantation might yield unfavor-able results in another site.

Degeneration of Biomaterial-Tissue InterfaceAs noted earlier, the wound healing response initially estab-lishes the tissue characteristics of the implant-tissue interface.Several agents can initiate degenerative changes in the interfacetissue. Other agents probably act as promoters to stimulate theproduction of proinflammatory mediators that stimulate tissuedegradation and potentiate the failure process.

Of the many factors affecting the implant-tissue interface,two of the most important are motion of the prosthetic compo-nent and particulate debris. It is difficult, however, to deter-

mine the causal relationships between these factors and implantfailure from studying only the end-stage tissue. Other histo-pathologic findings and laboratory studies indicate that metal

ions and immune reactions might play roles in the degenerativeprocesses leading to prosthesis loosening in certain patients.Systemic diseases and drugs used to treat the disorders couldalso contribute to the breakdown of the implant-bone interface.


Finally, interindividual differences in genetically determinedcellular responses might explain why prostheses fail in somepatients with a low mechanical risk factor for failure.

EFFECTS OF IMPLANT MOVEMENT. Movement of the

implant relative to the surrounding tissue can interfere with thewound healing response by disrupting the granulation tissue.With bone implants this relative movement, if excessive, candestroy the stroma required for osseous regeneration, and afibrous scar results.

Another important effect of implant motion is the forma-tion of a bursa within connective tissue, in which shearing andtensile movement has led to disruption of tissue continuity andformation of a void space or sac (lined by synovial-like cells).Therefore the tissue around prosthetic components that areremoved because of loosening might display features ofsynovial-like tissue. The presence of synovial cells (macrophageand fibroblast-like cells) is important because they could beactivated by other agents, such as particulate debris, to produceproinflammatoty molecules. The process of activation of thistissue might be similar to that occurring in inflammatory jointsynovium or bursitis.

Previous studies explain how prosthetic motion leads to theformation of the synovial-like tissue by showing that "synoviallining is simply an accretion of macrophages and fibroblastsstimulated by mechanical cavitation of connective tissue.,,36

These findings are based on experiments in which the mechan-ical disruption of connective tissue was produced by injectionof air and fluid into the subcutaneous space of animals.94 Theresulting sac was initially described as a "granuloma pouch."Later studies demonstrated that the membrane lining thepouch displayed the characteristics of synovium, and this tissuewas referred to as "facsimile synovium."36

Prosthetic motion can also contribute to wear of the pros-thetic component abrading against the bone cement sheath orsurrounding bone. This generates increased amounts of partic-ulate debris, which might contribute to activation of the mac-rophages and synovial-like cells at the implant-tissue interface.

EFFECTS OF IMPLANT-DERIVED PARTICLES. Particu-

late debris can be generated from the abrasion of the implantagainst surrounding tissue. The potential for wear is greaterwith materials or devices rubbing against a hard surface such asbone and with the articulating components of joint replace-ment prostheses. This particulate debris can induce changes inthe tissue around the implants. Adverse responses have beenfound to both metallic and polymeric particles.

The biologic reactions to particles are related to (1) particlesize, (2) quantity, (3) chemistry, (4) topography, and (5) shape.Although the role that each of these factors play in the biologicresponse is unclear, particle size appears to be particularlyimportant. Particles small enough to be phagocytosed (lessthan 10 11m) elicit more of an adverse cellular response thanlarger particles.

Particulate metallic particles (e.g., cobalt-chromium alloy)can induce rapid proliferation of macrophages and focaldegeneration of synovial tissues.49 Animal investigations and

histopathologic studies of tissues from human subjects havesuggested that titanium alloy is more "biocompatible" thancobalt-chromium alloys. Therefore researchers assumed thattitanium particulate debris would be less problematic thanparticles of cobalt-chromium alloy. Histology of pigmentedtissue surrounding titanium implants has generally revealedconsiderably fewer macrophages and multinucleated foreignbody giant cells than around cobalt-chromium alloy particlesand polymeric particulate debris. However, titanium alloyparticles generated by the abrasion of femoral stems againstbone cement in human subjects can cause histiocytic andlymphoplasmacytic reactions to the metallic particles. 1 Tita-nium particles have also been found to cause fibroblasts inculture to produce elevated levels of prostaglandin £2 (PGE2).These findings show that adverse effects may occur with thebiologic response to titanium particles as well as to cobalt-chromium alloy particulate debris.

Many investigations evaluating the histologic response topolyethylene and PMMA particles in animals and in tissuerecovered from revision surgery have revealed the histiocyticresponse to these polymer particles. This macrophage responsecan also lead to bone resorption.

Synovial cells also respond to calcium-containing ceramicparticles.79 Local leukocyte influx, proteinase, PGE2, andtumor necrosis factor (TNF) levels have been measured after

injection of calcium-containing ceramic materials into the"air-pouch model" described earlier. TNF was detected insignificant amounts after injection of the ceramics. Thesesubstances also caused elevated leukocyte counts and increasedlevels of proteinase and PGE2. Substances with surface chemis-tries that elicit a beneficial tissue response (e.g., bone bonding)when implanted in bulk form can cause destructive cellularreactions when present in particulate form.

Investigations indicate that most biomaterials, when pre-sent in particulate form in a size range small enough to bephagocytosed (less than 10 11m), can elicit a biologic responsethat could cause the bone resorption that initiates and promotesthe loosening process. This degenerative process has been re-ferred to as small-particle disease.

METALLIC IONS. Animal and human investigations haverevealed elevated levels of metal ions in subjects with certaintypes of implants (e.g., total joint replacement prostheses). Themechanisms of metal ion release are still being studied, andresults often vary when determining the concentration ofspecific metal ions in tissues and fluids. Metal in ionic form isoften not distinguished from that present as particles, whichconfounds interpretation of results.

Serum and urinary chromium levels increase in patientswho have undergone conventional cemented cobalt-chromiumalloy hip replacement.6 An attempt to determine the valency ofchromium as either +3 (III) or +6 (VI) from the metal ion

concentration in a blood clot was not successful. This experi-ment was based on erythrocytes displaying a unidirectionaluptake of Cr (VI) while effectively excluding Cr (III). Thedistinction of the valency of chromium is important because Cr(VI) is much more biologically active than Cr (III).


Unfortunately, our knowledge of the local and systemicbiologic and clinical sequelae of metal ion release has notsignificantly advanced over the past several years. Additionof cobalt ions in the form of cobalt fluoride solutions to the

media of synovial cells can stimulate their production ofneutral proteinases and collagenase.37 These findings may berelevant to tissue degradation (e.g., osteolysis) around im-plants, since metal ions could activate synovial cells in thesurrounding synovial tissue to produce agents that promotetissue degeneration.

DISEASES AND DRUGS. Implant failure has not been wellcorrelated with disease states and drugs used to treat thedisorders. Some observations indicate that antiinflammatoryagents, as well as certain anticancer drugs, can reduce theamount of bone formation around devices in the early stages ofwound healing after implantation. Little is known, however,about the role of these and other agents on tissue remodelingand degeneration at the biomaterial-tissue interface.

Systemic Response to Particulate DebrisLocal and regional lymphadenopathy caused by wear particlesreleased from implants is becoming increasingly recognizedas a possible complication with certain prostheses. Particlesgenerated by mechanical wear of prostheses can leave the site ofthe implant via the lymphatics and become engulfed bymacrophages within local and regional lymph nodes. Accumu-lation of cells containing particles causes enlargement of thelymph node and the characteristic histologic appearance ofsinus histiocytosis.44

Distension and prominence of the lymphatic sinuses resultfrom large numbers of (1) histiocytes derived from the cells thatline the sinuses or (2) macrophages derived from circulatingmonocyres. Multinucleated giant cells, resulting from the fu-sion of macrophages or histiocytes, may also be found in thedilated sinuses.

Accumulation of polyethylene, PMMA, and metal particlesin lymph nodes draining joints replaced with prostheses hasbeen found in animal69.112and human studies.7.14.47.57Some

have reported lymphadenopathy in surgical patients.44.86.95With joint replacement prostheses, synovial macrophages

readilyengulf particles released into the joint space. When theproduction of particulate debris exceeds the phagocytic capac-ity of synovial macrophages, excess particles enter lymphaticvessels.1l4 Evidence indicates that macrophages laden withparticles can also gain entry to the lymphatics.46 Macrophagespresent within lymph nodes endocytose free particles travelingwithin the lymphatic system. A steady influx of wear debriscausesthese macrophages to accumulate within the sinus of thelymph node.14 Over several years, macrophages with particlesmay become so abundant that they cause dilation of nodalsinusesand nodal enlargement. As mentioned, accumulation ofhistiocyres or macrophages within lymph node sinuses is de-scribedpathologically as sinus histiocytosis.

SYSTEMICMIGRATION OF PARTICLES. Numerous stud-

iesreport the migration of particles, released from implants, to

lymph nodes and many organs. The spread of particles fromsilicone elastomer and liquid droplets (e.g., from breastimplants) is well documented.104 The translocation of theseparticles results from the following:

1. Migration through soft tissues2. Entry into the lymphatic system3. Direct entry into the vascular system.Silicone particles have been found to migrate from breast

implants through soft tissue to sites as distant as the groin.20The finding of silicone lymphadenopathy in axillary lymphnodes is common in patients with breast implants.105 Thehematogenous dissemination of silicone to viscera has also beenreported as a result of soft tissue injection of the material. 104Inthe orthopedic literature, silicone lymphadenopathy has be-come a common finding in patients receiving finger jointprostheses made of silicone elastomer.24

Reports documenting dissemination of particles in thelymphatic system from total joint prostheses are mounting,suggesting that this phenomenon may be more commonthan previously thought. Several animal studies have docu-mented lymphatic spread of polyethylene particles to regionalnodes.69.112 Bos et al14 recently provided evidence fromhuman autopsies that polyethylene, PMMA, and metalparticles released from stable total hip replacements spread toinguinal, parailiac, and paraaortic lymph nodes as early as1112years after implantation of the prosthesis.

Sinus histiocytosis in association with wear particles ofpolyethylene has been an incidental finding in lymph nodesbiopsied at revision arthroplasri9 and in the staging of pros-tate7.44 and breast cancer.86 Adenopathy related to an implantis not limited to total hip and knee replacement prostheses.Axillary histiocytic lymphadenopathy was reported in associa-tion with polyethylene wear particles from a total shoulderreplacement.86

CliNICAL IMPliCATIONS. Lymphadenopathy secondaryto the accumulation of wear particles in sinus macrophagesmay make the appropriate diagnosis difficult, especially whenmalignancy is suspected. For example, a 19-year-old man hadright inguinal pain and a 3-cm2 palpable mass 3 years aftera right total knee replacement following resection of anosteosarcoma.95 The lymph node was biopsied to evaluate forsuspected metastatic recurrence of osteosarcoma. Histologicexamination revealed sinus histiocytosis caused by metal par- 'ticles released from the knee prosthesis. The patient had noevidence of malignancy.

The ultimate fate of particles released from total jointprostheses is unknown. A recent report suggests that metallicparticles from orthopedic prostheses may pass through thelymphatics and gain a systemic distribution. 57

Immune Reactions and Genetic Determinants

Two patients matched for gender, age, weight, activity level,and other factors that might affect prosthesis performanceoften have very different outcomes. This suggests that immunereactions or genetically determined responses might playa rolein the failure of prostheses in some patients.


Immune responses include antibody and cell-mediatedreactions and activation of the complement system. Certainmetal ions can behave as haptens, which can trigger animmune response when complexed with serum proteins. Thecell types that might be expected to occur at sites of antibodyand cell-mediated reactions, however, are not otten found in

tissue retrieved. These cells include lymphocytes and plasmacells. The finding of occasional lymphocytic infiltrates in tissuearound implants does not provide enough information.Immune reactions to polymeric materials (e.g., silicone) havealso been suggested as the cause of certain systemic diseases,but mechanisms for such a response, as well as its prevalence,remain in question. Much more work is necessaty to determinethe role of immune reactions in the response to implantabledevices.

Previous studies have demonstrated that many biomateri-als can activate (cleave) certain molecules (C3 and C5) in the

complement system and thereby stimulate the alternative path-way of the immune response. Complement activation by bio-materials may playa role in adverse reactions to certain devices.Again, however, additional studies are required.

One form of cell-mediated immune reaction associated

with implants that has been studied is the delayed hypersensitiv-ity response."Metal allergy" has been incriminated as the causeof failure in certain patients,72 but results are not yet definitive."The incidence of metal sensitivity in the normal population is

nlgn, witn up to 1'5% o{tne population sensitive to nicKel'and'perhaps up to 25% sensitive to at least one of the common

sensitizers Ni, Co, and Cr. The incidence of metal sensitivityreactions requiring premature removal of an orthopedic deviceis probably small (less than the incidence of infection). Clearlythere are factors not yet understood that caused one patient butnot another to react. "71

A similar situation exists with respect to sensitivity reactionsto polymeric materials, including bone cement (PMMA). Themonomer of PMMA is a strong skin sensitizer.7o Failure ofcemented devices, however, has not yet been correlated with ahypersensitivity response in patients.

No clear etiology exists for prosthesis loosening in somepatients, and in others with multiple risk factors for failure theprosthesis functions well. Therefore some suggest that geneticdeterminants may be responsible for loosening.

One investigation has shown interindividual differences in

the in vitro cytokine and PGEz production by lipopolysac-charide-stimulated macrophages.77 HLA-DRz-positive indi-viduals and first-degree relatives were found to be low respond-ers. Such studies suggest that in certain individuals, geneticdetermination might affect the degree to which cells in thetissue around prostheses can be activated to produce proinfIam-matory agents that stimulate tissue degradation. Additionalstudies could help identify patients who might be high re-sponders to prosthetic motion and particulate debris and there-fore at high risk for prosthesis failure.

CarcinogenicityChromium and nickel are known carcinogens, and cobalt is asuspected carcinogen. Therefore the release of these metal ionsinto the human body from implants might raise some concern.

Fortunately, reports of neoplasms around implanted devices(e.g., total joint replacement prostheses) have been few.Although no causal relationship has been shown, the index ofsuspicion is high enough to warrant serious investigationthrough epidemiologic and other studies. The use of porouscoated metallic devices (e.g., noncemented total joint replace-ment prostheses with large surface area) in younger patients hasadded to the concern aboUt the long-term clinical conse-quences of metal ion release.

The relationship of metallic ion release to oncogenesis1z andreports of neoplasms around orthopedic implants have beenreviewed.68 Differences in the tumor types, time to appearance,and type of prosthesis confound attempts to associate theneoplasm with the implant materials and released moieties.

In an epidemiologic investigation in New Zealand, morethan 1300 total joint replacement patients were followed todetermine the incidence of tumors at remote sites.43The inci-

dence of neoplasms in the lymphatic and hemopoietic systemswas found to be significantly greater than expected in thedecade after arthroplasty. Importantly, the incidence of cancerof the breast, colon, and rectum was significantly less thanexpected. The investigators acknowledged that although theassociation might be an effect of the prosthetic implants, othermechanisms, particularly drug therapy, require consideration.

Somewhat similar results were obtained from another studyof the cancer incidence in 443 patients who had total hiprepl'acement (McKee-Farrar) oerween 1967 ana 1973 ana were

followed through 1981. The risk ofleukemias and lymphomasincreased, whereas the risk of breast cancer decreased. Theauthors concluded that the local occurrence of cancer associ-

ated with Co-Cr-Mo prostheses indicates that "chrome-cobalt-alloy plays some role in cancerogenesis (Sic).,,109

Bacterial Infection

Under certain circumstances, biomaterial surfaces can providefavorable substrates for the colonization of bacteria. The adher-

ence of bacteria to solid surfaces is facilitated by their produc-tion of a biofilm. The biofilm is a complex structure comprisingbacterial cells encapsulated in a polymeric matrix. The detailedcomposition of the matrix has yet to be determined. Cettainbiomaterials may favor the production and adherence of abiofilm, and certain material characteristics may predispose tobacterial colonization. Studies are underway to understandthese processes.

BIOMATERIALFAILURE

Potential causes of failure of implanted devices include thefollowing:

1. Deficiencies in design of the device for a particularpatient

2. Surgical problems (e.g., problematic orientation andexcessive surgical trauma leading to problems in woundhealing)

3. Host abnormalities and diseases4. Infections

5. Biomaterial fracture, wear, and corrosion


In what cases has the failure of the biomaterial initiated thefailure of the device? It is difficult to determine whetherbreakdown of one or more biomaterials used in the fabrication

of a particular prosthesis was the primary cause of the systemfailure or secondary to other factors. Diagnostic imagingmethods such as radiography and computed tomographyrarely provide evidence of biomaterial failure early enoughfor it to be identified as the primary cause of implant failure.Recovery of the device and surrounding tissue at revisionsurgery provides limited information about the initiatingcauses of failure.

All the materials used to fabricate devices have finite "fatiguelives" (i.e., failure as a result of cyclic loading). Unfortunately,the endurance limit for many implantable devices or the mag-nitude of the applied stress is not known. How does the device'sserviceable life relate to the patient's longevity.

The introduction of the "supermetals" virtually eliminatedconcerns about breakage of metallic devices. Corrosion canoccur, but only rarely is it the initiating cause of device failure.Wear, however, is clearly the principal cause of failure incertain articulating devices (e.g., joint replacement prostheses)and can occur to some extent in soft tissue implants.

Breakdown of a biomaterial does not necessarily cause theimmediate failure of a device. In fact, biomaterial failure mightbe considered an ongoing process. It becomes clinicallyimportant when it is the initiating cause of failure or acceleratesa degenerative process initiated by other factors.

Wear

Wear is generally defined as the loss of material from solidsurfaces as a result of mechanical action. Several mechanisms

can contribute to wear. In each case, fracture through thesubstance (i.e., loss of cohesive bonding) must occur for afragment of the material to be removed from the surface. Thethree main mechanisms of wear are (1) adhesive wear,

(2) abrasive wear, and (3) fatigue wear.Adhesive wear occurs when the force of adhesion between

contacting surfaces exceeds the cohesive force within one ofthe materials. When two surfaces come into contact, atomic

bonding can occur between the materials (Figure 19-4 andTable 19-1). Generally, when the surfaces separate, this bond isbroken. In some cases, however, this bond strength exceeds the

Harder material

Softer materialCrack propagated byfatigue {delamination}wear

Figure 19-4. Processes in adhesive, abrasive. and fatigue(delamination)wear.

strength of the atomic bonding in one of the two contact-ing materials. When the two surfaces separate, fractureoccurs within the material, leading to loss of a transferredfragment.

In adhesive wear the amount of material lost is generallydirectly proportional to the applied load and to the distanceslid. In many material systems the amount of wear is inverselyproportional to the hardness of the surface being worn away.These relationships are incorporated in the following equation:

kWLV.dh= H

v..dhis volume of adhesive wear, and k is a dimensionless wear

coefficient related to the probability of particle formation, orthe fraction of adhesive junctions that eventually produceswear particles; k is generally less than 0.1 (i.e., only one of 10adhesive junctions results in a wear particle). W is the loadperpendicular to the surface; L is the distance through whichthe surfaces slid; and H is the indention hardness of the softermaterial.

Abrasive wear results when asperities on one material plowmaterial from a softer contacting surface (Figure 19-4). Theloss of material that occurs as two surfaces contact is referred

to as two-body wear. When a third substance is interposedbetween the two surfaces, the process is three-body wear. Thefollowing equation has been used to evaluate the volume ofmaterial lost in abrasive wear:

WL

V.br= H p tan e

Table 19-1.Wear Processes

TYPE MECHANISM PART SIZE

Adhesive Chemicaladhesion

nm to !!m

Abrasive (twobody)

Plowing ofasperitythroughsoftermaterial

Abrasive (threebody)

Entrapment andplowing ofparticle

Propagation ofsubsurfacecracks to

surface bycycliccompression.tension, orshear

!!m to mmFatigue(delamination)

252 PART I PRINCIPLESAND TECHNIQUES

v..br is the volume of abrasive wear, and e is one-half theincluded angle of the asperity of the harder material plowingthrough (i.e., abrading) the softer material. Tan e is very smallfor sharp aperities, leading to high wear volume.

Fatigue wear results from repeated loading of contactingsurfaces. The cyclic stresses initiate or propagate surface orsubsurface cracks (Figure 19-4). This process eventually leadsto the loss of relatively large fragments of material.

The term fretting wear is used to describe the loss of material&om contacting surfaces undergoing oscillatory tangentialdisplacement of small magnitude (micrometers). The loss ofmaterial from fretting wear is the result of adhesive or abra-sive wear.

Another form of wear involves the process of corrosion. Theoxide passivation layer on metals may be lost as a result of awear process. The removal of this film may accelerate thecorrosion process and lead to alterations in the oxide layer.These alterations may include an increase in rugosity (surfaceroughness), which can promote abrasive wear of the opposingsurface.

Corrosion

Whereas wear is the loss of solid fragments from surfaces as aresult of mechanical action, corrosion is release of ions and

compounds as a result of chemical action. Relatively little isknown about the mechanisms underlying the corrosion of

u'CItlMecan\\:;-sao'sram:es.-tfi'e r611~wlligalSCUSSll'mtocuses oncorrosion of metallic substances.

Although they are relatively inert, metals are soluble inaqueous solUtions. Metal leaves the solid metallic state to form

aqueous cations in electrolyte solutions (Mn+aqueousin follow-ing reaction). Passivation, the formation of an insoluble salt(oxide) on pure metals and metal alloys, inhibits metal egressand thus inhibits corrosion. This layer selVes to protect themetal by insulating it from the electrolyte solution. A chro-mium oxide passivation layer forms on stainless steel andcobalt-chromium alloy. A titanium oxide layer forms on tita-nium and titanium alloys.

The electrochemicalseries(or galvanic series) of metals ratesthe relative tendencies of metal ions to go into solution (Table19-2). This classification of metals, however, does not consider

the effect of an oxide layer on the metal's surface. The electro-chemical series shows that metallic elements such as gold arerelatively insoluble, whereas substances such as titanium andaluminum, when present in pure elemental form, are relativelysoluble (and active). When these same reactive metallic ele-

ments are covered by their oxide layers, however, these metalstend much less to go into aqueous solution.

Titanium is a more active element than chromium; titanium

and its alloys form oxide passivation layers more rapidly than dosubstances containing chromium (i.e., cobalt-chromium alloyand stainless steel). Although an active metal such as titaniumforms its oxide passivation layer spontaneously in any environ-ment containing oxygen, however, the strength of adhesion ofthe oxide layer to the underlying titanium metal is not as greatas that of the chromium oxide layer to its metal substrate. Inaddition, the chromium oxide passivation film is more densethan the titanium oxide layer.

The following reaction describes the ion transfer

Mn+aqueousinto solution (i.e., the corrosion) that occurs whena metal is immersed in an electrolyte solution. Electrons, freedas the metal ions enter solution from one region of the speci-men (the anode), travel to other regions of the same specimen,making those areas cathodic. If the metallic sample is connectedto (e.g., touching) another metal specimen with less of a ten-dency for corrosion, the second metallic specimen becomes thecathode (Figure 19-5). Corrosion of the anode can be acceler-ated by increasing the rate of reaction at the cathode.

Anodic reaction

M ~ Mn+ aqueow + ne- metal

Cathodic reactiom

Reduction of dissolved oxygen

n (I)

n

1: \2 O2 + 1:H20 + ne-metal ~ nOfr aqueous

Reduction of hydrogen ionsn

nW+ne- ~-H22

Table 19-2.Electrochemical Series of Metals with NormalElectrode Potentials*

*Measured in volts at 25° C (77° F). referred to hydrogen as zero.

METAL ELECTRODE POTENTIAL (V)

NOBLE ENDGold +1.45

Platinum + 1.20

Silver +0.80

Copper +0.34

Hydrogen 0.00

Molybdenum -0.20

Nickel -0.25

Cobalt -0.28

Iron -0.44

Chromium -0.73

Titanium -1.63

Aluminum -1.66

Magnesium -2.37

Lithium -3.05

ACTIVE END


, - -~'- ' 'I ~""'ofh;gh

A,,,,,of low 0 ~ ><yge"","","trot;O" 0oxygen concentration 0 0 0 0 0 0 0

~O ~00000008

8o0 000

. / . "-- (') 000 0, ' , n

The mechanisms controlling corrosion processes relate tofactors that favor either the anodic reaction or the cathodicreactions. Factors that assist electron transfer from the metal

would facilitate the anodic reaction, whereas changes in oxygenconcentration at sites along the metallic surface would facilitatethe cathodic reaction.

Activemetal

(anodic,oxidation)

++

00

00

~n How -0~1 Passivemetal(cathodic,reduction)

"0+

+ _00-

++ +

Electrolyte

Figure 19-5. "Concentration cell" depicting types of reactionsincorrosion of metals.

Depletion

Figure 19-6. Propensity of oxygen-deleted regions to corro-sionon metallicsurfaces.

Crevice

\ '\/

/

I Metal/'

\ Metal \ /

",/

The mechanisms of corrosion referred to as pitting, crevice,and depletion relate to situations in which oxygen is relativelydepleted at certain sites on a metal implant (Figure 19-6).These sites take on an anodic character, favoring metal ionrelease. Salts can accumulate at the sites of scratches producedby instruments at insertion. The depletion of oxygen at thesesites can lead to corrosion, resulting in pitting of the surface.The low oxygen concentration at the bottom of the pits favorscontinuing corrosion.

Crevices formed at the sites where components are joined(e.g., in modular components of prostheses or at screw-platejunctions) also create a microenvironment favorable to corro-sion as a result of depleted oxygen (Figure 19-7). The depletionof oxygen can also occur under plaques of biologic debris on thesurface of devices, thus also predisposing to a so-called concen-tration cell corrosion mechanism.

Another basic mechanism of corrosion can occur whendissimilar metals touch. The tendencies of electrons to movefrom the more active metal to the more noble metal causes the

former to become anodic, thereby producing conditions thatwould favor accelerated corrosion. This mechanism of corro-

sion has been referred to asgalvanic, two-metal, mixed-metal, orcouple corrosion. The degree of galvanic corrosion depends onthe electrochemical nature of the tWo metals and the relative

areas of contact and surfaces exposed to the solution.The term galvanic corrosion is derived from an observation

made by Galvani in 1791 as he was investigating the suscepti-bility of nerves to irritation. He found that if a rod of brasscontacted the frog's foot while a silver rod contacted the spinalcord, the leg muscles contracted when the free ends of the rodstoucned. In 1800 Vofta confirmed that the force that provokedthe contraction was electrical in nature. The electric current

generated by contact of dissimilar metals in an electrolytesolution is referred to as galvanic current.

In certain situations the breakdown of material is caused bymechanical and chemical processes acting in concert. Thesemechanochemical processes include (1) fretting corrosion,(2) stress corrosion, and (3) metallic transfer (with subsequentgalvanic corrosion).

Fretting corrosion is a process in which abrasive wear isaccompanied by corrosion. The protective oxide layer on the

"-

"-"

...."\

"- // '\

Figure 19-7. Crevices in which oxygen concentration is lower than that around metallic deviceare predisposed to corrosion.

I- /" - \ Anodic ),'- '" I"

" 1 -//

1 " I" /Metal - I // - " / - "I


metal is removed by abrasive wear. The new passivation layerthat forms after abrasion is not as durable or chemically inert asthe original layer, thereby making the metal more susceptibleto corrosion. Stainless steel and cobalt-chromium alloys aresusceptible to fretting corrosion. This form of corrosion oftenoccurs between screw heads and bone plates.

Stress corrosion is the process by which the presence of anelectrolyte decreases the strength of the metal. Microcracksdevelop at anodic areas of the metallic device as the result ofcorrosion processes. These cracks propagate under appliedstress. The tips of these microcracks are at a more highlystrained state than surrounding metal and are more susceptibleto corrosion, which contributes to crack propagation.

.ABdissimilar metals come in contact, metallic tramfer ofsmall fragments of the softer metal can occur to the harder one..ABthese two metals remain in contact, galvanic corrosionprocesses may develop. Therefore implants should be handledwith instruments made of the same metal as the device.

Several methods are available to assess the predisposition ofmetals to corrosion. A simple method involves weighing thespecimen after exposure to corrosive environments and assay-ing the bathing electrolyte solution for corrosion products. Anelectrochemical method that has been valuable in characteriz-

ing the potential of metals to corrode is anodicpolarization. Inthis experimental procedure the test metal is the anode of anelectrolytic cell. An increasing voltage is applied between it anda reference cathode (usually platinum). The current demity(current divided by cross-sectional area of the anode) is ameasure of the corrosion rate of the anodic metal in the partic-ular electrolyte used.

Table 19-2 shows the electrochemical potentials and ten-dency for corrosion (i.e., current density) for the orthopedicalloys. The current density for titanium alloy is significantly lessthan that of the cobalt-chromium alloy and stainless steel,indicating that titanium and its alloys have much less of atendency for corrosion. Because the release of metal depends onsurface area, a greater amount of metal is released from porousmetal specimens.

BIOMATERIALSFOR TISSUE ENGINEERING

The wound healing process in many types of lesions inconnective tissues does not lead to spontaneous regeneration;the end result is often a dysfunctional scar. Recent research hasshown that some cells can proliferate and maintain theirphenotype when cultured on certain two-dimensional sub-strates or in specific three-dimensional porous matrices andgels in vitro. This has formed the basis for tissue engineering,with the promise of forming tissues in vitro for subsequentimplantation.

Initially, "tissue engineering" was used principally to de-scribe tissue produced in culture by cells seeded in porousabsorbable matrices. 55 More recently, however, the scope oftissue engineering has been widened to include the implemen-tation of porous matrices, alone or seeded with cells, as implantsto facilitate tissue regeneration in vivo. This enfranchises de-

cades of investigation of wound healing and treatments imple-menting absorbable implants.

In Vivo Versus In Vitro StrategiesJust as they are the three components of tissue, the matrix,cells, and soluble regulators are primary elements in strategiesto engineer tissue in vivo or in vitro for subsequent implanta-tion. Decisions aboUt which elements might be required forregeneration of tissue in vivo can be guided by an understand-ing of the deficits of the natural (i.e., spontaneous) healingprocesses that prevent regeneration.

An advantage of tissue synthesis in vitro is the ready abilityto examine the material as it is formed and to make certain

measurements to establish its functions before implantation. Adisadvantage, particularly in the production of tissues thatmust playa load-bearing role, is the absence of a physiologicmechanical environment during the formation of the tissue invitro. Mechanical force serves as a critical regulator of cellfunction and can profoundly influence the architecture oftissue as it is forming.21.93 Because the mechanical environ-ment extant during formation of most load-bearing tissue invivo is not well understood, it is not yet possible to recreatesuch an environment in vitro during the engineering of mosttissues.

Another disadvantage in the formation of load-bearing tis-sue oUtside the body is the necessary incorporation of the tissuesafter implantation, which requires that the engineered tissue bemechanically coupled to the surrounding structures. Union ofthe implanted tissue with the host organ requires remodeling(degradation and new tissue formation) at the interfaces of theimplant with the host tissues. Remodeling of implanted tissueengineered in vitro is essential for its functional incorporation.This demonstrates the benefit of de novo regeneration in vivo,with the incorporation occurring as the tissue is being formed.

Thus, for certain tissues, an effective strategy may be tofacilitate tissue formation in vivo, under the influence of the

physiologic mechanical environment. One disadvantage ofthis approach, however, is that the regenerating tissue may bedislodged or degraded by the mechanical forces normallyacting at the site, before it is fully formed and incorporated.

Matrices

Matrices for engineering bone and soft connective tissues haveincluded synthetic and natural calcium phosphates51.60 andmany synthetic (e.g., polylactic acid,25.76 PGA,40.56) andnatural polymers (e.g., collagen,81.102 fibrin48,52). Materialused in the fabrication of matrices for engineering tissue invitro or used as implants to facilitate generation in vivo musthave the necessary microstructure and chemical compositionto accommodate parenchymal cells and their functions. In thisregard a porous structure is generally needed. The requiredpercentage of porosity and the pore diameter, distribution, andorientation may vary with tissue type. Moreover, because theobjective is the regeneration of the original tissue, the scaffoldneeds to be absorbable.

The chemical composition of the matrix is important withrespect to its influence on cell adhesion and the phenotypic


expression of the infiltrating cells. The degradation rate of thematerial generally may be determined based on the rate of newtissue formation and the normal period for tissue remodelingat the implantation site. It is important to consider the effectsof moieties released during matrix degradation on the host andregenerating tissue.

A matrix can play the following roles during the process ofregeneration in vivo:

1. The matrix can structurally reinforce the defect site tomaintain the shape of the defect and prevent distortionof surrounding tissue. For example, cysts that form inthe subchondral bone underlying the articulating sur-faces of joints can lead to collapse of the joint surface.

2. The matrix can serve as a barrier to the ingress ofsurrounding tissue that may impede the process ofregeneration. The concept of guided tissue regenerationis based in part on the prevention of overlying gingivaltissue from collapsing into the periodontal defect.32

3. The matrix can serve as a scaffold for migration andproliferation of cells in vivo or for cells seeded in vitro.

4. The matrix can serve as an insoluble regulator of cellfunction through its interaction with certain integrinsand other cell receptors.

One class of properties that is likely to play an especiallyimportant role in the engineering of load-bearing tissues ismechanical properties. Design specifications for several bonegraft substitute materials have included the strength of thematrix material, with the principal objective of employinghigh-strength substances for immediate load bearing. Syn-thetic calcium phosphate ceramics were implemented as matrixmaterials to facilitate bone regeneration in vivo. 17.33

Besides possessing high strength, however, some of thesematerials also have a high modulus of elasticity, making themextremely stiff structures. The disadvantage of this extremestiffnessis that the presence of the material greatly alters thedistribution of mechanical forces in surrounding tissue andthus adversely affects the stress-induced remodeling of neigh-boring bone. Because many materials of this class of matrix areessentially nonresorbable, the adverse effects on remodelingwill persist indefinitely. This remodeling can result in os-teopenic regions around the implanted site, increasing the riskof fracture. Further complications can arise because conven-tional methods for treatment of such fractures are likely to bedifficult, since the high density of the material precludes dril-ling and sawing procedures. Revision surgical procedures atsites implanted with these substances may require their com-pleteremoval.

Based on these considerations, the materials used should

match the modulus of elasticity of the tissue at the implant site.Thus the makeup and properties of the extracellular matrix(ECM) should be considered in selecting or designing ascaffoldfor tissue engineering.

One design approach has been to employ matrices thatserveas analogs of the ECM of the tissue to be engineered. 119This concept recognizes that the molecular composition andarchitecture of the ECM display chemical and mechanicalproperties required by the parenchymal cells and the physio-

logic demands of the tissue. For scaffolds in regeneration ofbone, this approach has led to the use of natural bone mineralproduced by removing the organic matter of bovine bone. *The calcium-deficient carbonated apatite, which constitutesthe mineral phase of bone, and the unique microstructureof the ECM of bone are determined by the organic templateof collagen and the initial release of calcium phosphate-containing vesicles from osteoblasts. This largely explains whybone mineral cannot be replicated in the laboratory.

Several synthetic and natural polymers have been used inthe development of implants for the treatment oflesions in softconnective tissues. For example, with cartilage defects, PGAand PGA-polylactic acid blends30.106 and natural polymers,including fibrin,48 collagen gel110and sponge,102 and hyaluro-nan, 58have been seeded with chondrocytes in studies in vitroand in vivo.

The approach of using analogs of ECM as implants for theregeneration of the soft connective tissue and nerve hasemployed porous collagen-glycosaminoglycan (GAG) copoly-mers. Regeneration of dermis in animals and human sub-jectsI8.115.116.118and reconnection ofaxons of cells in ruptured. h I . 23 II6 120 . .. .fipenp era nerves m rats' . reqUire certam tIssue-specl cpore characteristics and degradation rate. To produce thesematerials, collagen is precipitated from acid dispersion in thepresence of chondroitin-6-sulfate.II9 The suspended copre-cipitate suspension is injected into a silicone tube (3.8 mminside diameter), or it is spread on a pan for immersion into acoolant bath and freeze-dried, to produce a porous architec-ture. The matrices are then exposed to a dehydrothermaltreatment or to ultraviolet light for cross-linking and steriliza-tion. Additional cross-linking can be achieved using aldehydesor other cross-linking agents. These collagen-GAG matricesare nominally 95% porous and can be produced with anaverage pore diameter of 30 to 120 ~m.

CALCIUM PHOSPHATE AND BONE GRAFf SUBSTI-

TUTE MATERIALS. Although some polymers have been in-vestigated as matrices to be used as implants to facilitate boneregeneration, most materials comprise calcium-containing sub-stances. Calcium sulfate88 and tricalcium phosphateI9.31.78.84were two of the calcium-containing substances first imple-mented as bone graft substitute materials. These materialsundergo physiochemical dissolution relatively quickly, disap-pearing in days to weeks in some cases.

Some question remains, however, about the rate at whichthese materials become absorbed by the body. Differences maybe related to variations in the composition and structure andthe different physiologic characteristics of the implant sites andanimal models. Tricalcium phosphate often undergoes physi-cochemical dissolution at a rate that precludes the precipitationof biologic apatite and subsequent bone formation on itssurface. The dissolving surface does not allow for proteinadsorption and cell attachment.

Work with hydroxyapatite began in the mid-1970s.34.50Many studies focused on the use of hydroxyapatite in dense

'References 10. 11.26.51.53.75.83.98.113.


and porous forms of block and panicles.45 The increasinginterest in bone graft substitute materials has advanced ourunderstanding of the bone response to these substances.Mechanisms underlying the bone bonding associated with in-corporation of these substances in osseous tissue are beginningto be revealed (see earlier discussion).

The mechanism of breakdown of calcium phosphates invivo is not fully understood. The appearance of macrophagesand multinucleated foreign body giant cells around some typesof calcium phosphates suggest that particles might provoke theactivation of phagocytes, which in turn might stimulate othercells and an inflammatory response. Agents produced by thesecells could accelerate the degradation process.

Although generally considered a nonresorbable material,synthetic hydroxyapatite has also been found to undergo phys-icochemical dissolution, albeit at a very slow rate. Because it isonly slightly soluble in biologic fluid, synthetic hydroxyapatitesubstances can functionally be considered as long-lasting im-plants, especially when they are incorporated into bone.

CARTILAGE ENGINEERING. Natural and synthetic materi-als have been employed as scaffolds for implantation in defectsof cartilage. With articular cartilage defects, matrices have beenused alone or seeded with cells before implantation. Nonre-sorbable materials that have been studied include carbon fiber

mesh15.74 and spongelike constructs of polytetrafluoroethyl-ene2.73and polyester.73

Concern about the effects of the permanent scaffold on themechanical performance of regenerated cartilage has focusedon resorbable substances as matrices. These include fibrin,48.87

collagen gels,111 collagen sponges,9.41,42.85.102PGA,39.97.107and polylactic acid.25.30.97

Questions remain as to the most suitable chemical compo-sition and pore structure in the fabrication of a cell-seededimplant to treat defects in the anicular surface. The substanceshould have a composition that maintains the chondrocytephenotype and a pore structure that accommodates cell infiltra-tion. Furthermore, the scaffold must be mechanically stableenough to be surgically manipulated for implantation. Becauseof the poor surgical handling characteristics of gels, spongelikematrices may be preferred.

Insoluble type I collagen matrices have been successful inanimal models to improve healing of cartilage defects.9.85.102Complete regeneration was not achieved, however, and thereparative tissue consisted of fibrous and fibrocartilaginoussubstances, with some hyaline cartilage. Recent work hasinvestigated the behavior of adult canine anicular chondro-cytes in vitro in matrices comprising collagen-GAG copoly-mers made from types I and II collagen.82

PERIPHERAL NERVE REGENERATION. One focus of ex-

perimental efforts to promote and improve peripheral nerveregeneration across gaps has been guiding the axonal growthcone to the distal stump. The proximal and distal nerve ends areencased in a tubel03-a nerve guide or nerve channel-in aprocedure referred to as "fascicular tubulization"28.91 and "en-

tubulation repair.,,67 This approach is analogous to strategies

employed to facilitate regeneration of certain connective tis-sues, including bone,8° periodontium,32 and tendon.64 Thelesion is isolated from the negative influence of surroundingtissues, particularly as they collapse into the defect, whilewound-derived trophic factors are maintained in the lesion site.With nerve the tube also serves as a guide for axonal elongation.Many synthetic and natural resorbable tubes, as well as silicone,have been investigated for this application. 8

Silicone tubulization of nerve gaps has been the standardexperimental model for many years59.65and has been shown toimprove regeneration compared with defects that were notcontained with a tube. In clinical use, however, the silicone

tubes have presented problems for long-term recovery. Siliconetubes typically have become encapsulated with fibrous tissue,which has led to constriction of the nerve, necessitating surgeryto remove the tube.

Several previous studies have investigated the performanceof biodegradable tubes fabricated from type I collagen.Archibald et al5 obtained the most extensive longitudinal datafor nonhuman primate peripheral nerve repair. Their 31h-yearstudy demonstrated that repair of a 5-mm gap in the monkeymedian nerve with a collagen tube had similar results toautograft repair.

Despite the promising results obtained using tubular nerveguides, incomplete regeneration and the desire to accelerate theprocess have prompted the investigation of soluble factors andinsoluble regulators (matrices) placed within the tubes beforeimplantation.8 It has been hypothesized that these substancesmay (1) provide more specific directional orientation for theaxons, (2) enable haptotaxis, and (3) act as growth and trophicfactors.38 Materials investigated as promoters of nerve regener-ation in tubes include laminin,8.67 gelatin with ACTH4-9,90type I collagen precipitate,66.92 fibronectin,66 and skeletalmuscle basal lamina grafts.16 The results of these efforts havebeen mixed; some substances did improve regeneration,whereas others actually impeded it. A recent study reportedresults of a collagen-GAG-filled tube that were comparable toan autograft in a rat model. 22

Future of Biomaterials for Tissue EngineeringTissue engineering may solve a number of compelling clinicalproblems in dentistry not adequately addressed using perma-nent replacement devices. The challenge will be to select theoptimal combination of matrix, cells, and soluble regulators fora particular clinical problem.

For many connective tissues of the musculoskeletal system,with microstructures that reflect the mechanical environment,

it may be more advantageous to regenerate the tissue in vivothan to engineer the tissue fully in vitro for subsequent implan-tation. The porous material that will serve as the matrix tofacilitate this regeneration must have certain pore characteris-tics, chemical constituents, and mechanical properties. Oneapproach has been to employ substances that serve as analogs ofthe ECM for the tissue to be regenerated.

With bone, natural bone mineral (anorganic bone) has beenefficacious in several experimental animal and clinical studies.For selected indications in which the supply of endogenous


precursor cells has been compromised by disease or priorsurgical procedures, it may be necessary to seed the matrixbefore implantation with exogenous cells or to use the matrixas a delivery vehicle for growth or differentiation factors.

REFERENCES

1. Agins HJ, Alcock NW, Bansal M, et al: Metallic wear in failed

titanium-alloy total hip replacements, ] Bone Joint Surg70A:347-356, 1988.

2. Ahfeld SK, Larson RL, Collins HR: Anterior cruciate

reconstruction in the chronically unstable knee using an

expanded polytetrafluoroethylene (PTFE) prosthetic ligament,

Am] Sports Med 15:326-330,1987.

3. Anderson JM, Miller KM: Biomaterial biocompatibility and

the macrophage, Biomaterials 5:5-10, 1984.

4. Andriano KP, Pohjonen T, Tormala P: Processing and

characterization of absorbable polylactide polymers for use in

surgical implants,] Appl Biomat 5:133-140, 1994.

5. Archibald SJ, Shefner J, Krarup C, Madison RD: Monkey

median nerve repaired by nerve graft or collagen nerve guide

tube,] Neurosci 15:4109-4123, 1995.

6. Bartolozzi A, Black J: Chromium concentrations in serum,

blood clot and urine from patients following total hip

arthroplasty, Biomaterials 6:2-8, 1985.

7. Bauer TW, Saltarelli M, McMahon JT, Wilde AH,: Regional

dissemination of wear debris from a total knee prosthesis: a

case report,] BoneJoint Surg 75A: 106-111, 1993.

8. Bellamkonda R, Aebischer P: Tissue engineering in the

nervous system. In Bronzino JD, editor: Biomedical engineeringhandbook, Boca Raton, Fla, 1995, CRC Press.

9. Ben-Yishay A, Grande DA, Schwartz RE, et al: Repair of

articular cartilage defects with collagen-chondrocyte allografts,

Tiss Engr 1:119-133, 1995.

10. Benezra V, Hobbs LW, Spector M: Ultrastructure and

architecture of bone mineral and synthetic bone substitutes

revealed by low-voltage high-resolution SEM. Presented at the

23rd Annual Meeting of the Society for Biomaterials, NewOrleans, 1997.

11. Berglundh T, Lindhe J: Healing around implants placed in

bone defects treated with Bio-Oss, Clin Oral Implants Res8:117-124,1997.

12. Black J: Metallic ion release and its relationship to oncogene-

sis. In Fitzgerald RHJ, editor: The Hip, Proceedings of the

Thirteenth Open Scientific Meeting of the Hip Society, St Louis,1985, Mosby.

13. Bobyn JD, Spector M: Polyethylene. In Encyclopedia of materi-

als, science,and engineering, New York, 1987, Pergamon.

14. Bos I, Johannisson R, L6hrs U, et aI: Comparative investiga-

tions of regional lymph nodes and pseudocapsules after

implantation of joint endoprostheses, Path Res Pract 186:707-716, 1990.

15. Brittberg M, Faxen E, Peterson L: Carbon fiber scaffolds in the

treatment of early knee osteoarthritis, Clin Orthop 307:155-164, 1994.

16. Btyan DJ, Miller RA, Costas PD, et aI: Immunocytochemistry

of skeletal muscle basal lamina grafts in nerve regeneration,

Pfast Recomtr Surg 92:927-940, 1993.

17. Bucholz RW, Carlton A, Holmes RE: Hydroxyapatite and

tricalcium phosphate bone graft substitutes, Orthop Clin NorthAm 30:49-67,1987.

18. Burke JF, Yannas IV, Quinby WC, et al: Successful use of

a physiologically acceptable artificial skin in the treatment of

extensive burn injury, Ann Surg 194:413-428, 1981.19. Cameron HU, Macnab I, Pilliar RM: Evaluation of a

biodegradable ceramic,] Biomed Mater Res 11:179-186, 1977.

20. Capozzi A, DuBou R, Pennisi VR: Distant migration of

silicone gel from a rupture breast implant, Pfast Reconstr Surg62:302, 1978.

21. Career DR, Orr TE: Skeletal development and bone fimc-

tional adaptation,] Bone Miner Res 7:389-395, 1992.

22. Chamberlain LJ, Yannas IV, Hsu H-P, et al: Collagen-GAG

substrate enhances the quality of nerve regeneration through

collagen tubes up to the level of autograft, Exp Neurol154:315-329, 1998.

23. Chang AS, Yannas IV, Perutz S, et aI: Electrophysiological

study of recovery of peripheral nerves regenerated by a

collagen-glycosaminoglycan copolymer matrix. In Gebelein

CG, editor: Progressin biomedical polymers, New York, 1990,Plenum.

24. Christie AJ, Weinberger KA, Dietrich M: Silicone lymphade-

nopathy and synovitis: complications of silicone elastomer

finger joint prostheses,]AMA 237:1463-1464, 1977.

25. Chu CR, Coutts RD, Yoshioka M, et aI: Articular cartilage

repair using allogeneic perichondrocyte-seeded biodegradable

porous polylactic acid (PLA): a tissue-engineering study,] BiomedMater Res 29:1147-1154,1995.

26. Clergeau L, Danan M, Clergeau-Guerithault S, Brion M:

Healing response to anorganic bone implantation in periodon-

tal intrabony defects in dogs. Part I. Bone regeneration: a

microradiographic study,] Periodontol67: 140-149, 1996.

27. Coleman DL, King RN, Andrade JD: The foreign body

reaction: a chronic inflammatory response,] Biomed Mater Res8:199-211,1974.

28. Colin W, Donoff RB: Nerve regeneration through collagentubes,] Dent Res 63:987-993, 1984.

29. Corrin B: Silicone lymphadenopathy,] Clin PathoI35:901-902, 1982.

30. Coutts R, Yoshioka M, Amiel D, et aI: Cartilage repair using

a porous polylactic acid matrix with allogenic perichondrial

cells. Presented at the 40th Annual Orthopedic Research

Society Meeting, New Orleans, 1994.

31. Cutright DE, Bhaskar SN, Brady JM, et al: Reaction of bone

to tricalcium phosphate ceramic pellets, Oral Surg 33:850-856, 1972.

32. Dahlin C, Linde A: Bone formation by guided tissue

regeneration. In Glimcher MJ, Lian JB, editors: The chemistry

and biologyof mineralized tissue, New York, 1988, Gordon andBreach Science.

33. Damien q, ParsonsJR: Bone graft and bone graft substitutes:a reviewof current technologyand applications,] Appl Biomat2:187-208,1991.

Bio Materials

Documents

Transcript of Bio Materials