Journal of Dental Research

http://jdr.sagepub.com/

Development and Evaluation of Porous Dental Implants in Miniature Swine

M.T. Karagianes, R.E. Westerman, J.J. Rasmussen and A.M. Lodmell J DENT RES 1976 55: 85 DOI: 10.1177/00220345760550013001 The online version of this article can be found at: http://jdr.sagepub.com/content/55/1/85

Published by:
http://www.sagepublications.com

On behalf of:
International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at: Email Alerts: http://jdr.sagepub.com/cgi/alerts Subscriptions: http://jdr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jdr.sagepub.com/content/55/1/85.refs.html

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Development and Evaluation of Porous Dental Implants in Miniature Swine

M. T. KARAGIANES, R. E. WESTERMAN, J. J. RASMUSSEN, and A. M. LODMELL Biology Department, Battelle Pacific Northwest Laboratories, Richland, Washington 99352, USA, and Walla Walla, Washington 99362
Porous metallic and ceramic dental implants were designed, fabricated, and implanted into fresh and healed alveolar sites of extracted mandibular premolar teeth in miniature swine. Bone ingrowth securely anchored the implants; intraoral devices were later attached to study the effects of stress on im-

plant stability.
The replacement of extracted teeth by endosteal implantation of prosthetic teeth is receiving increased attention. The Journal of the American Dental Association has published an extensive review listing 274 references on the current status of such implants.' Basically, most techniques' used must be viewed as simply a penetration of bone by a foreign object, often followed by soft tissue encapsulation. However, we and others2-4 have shown that bone will infiltrate suitable inert porous materials, yielding a good boneprosthesis bond. This study has as its major long-term goal the development of easily fabricated and implantable porous endosteal implants. Our short-term objectives included investigating the suitability of specific porous titanium and ceramic materials for implant applications, developing satisfactory designs and fabrication procedures for producing these
devices, and defining bone-implant interface bonding characteristics through in vivo animal studies.

Materials and Methods The principal metallic material studied is

This investigation was supported by Contract No. NIH-71-2386 from the National Institute of Dental Research, National Institutes of Health, Bethesda, Md. Received for publication November 4, 1974. Accepted for publication July 28, 1975.

void-metal composite (VMC), developed at this laboratory. VMC consists of a Ti-6A14V alloy with interconnecting and controlled porosity achieved through special fabrication techniques.5 The voids can be varied precisely in shape (for example, spheres or cylinders) or orientation to alter the mechanical and physical properties of the metal as well as the structure of the metal-tissue interface. The ultimate compressive strength of 50% dense VMC (that is, a density of 0.5 times that of the solid alloy) is -30,000 psi. The modulus of elasticity of the spherical-pore VMC is 5 (+ 1) X 105 psi, and that of the cylindrical-pore VMC is 9.5 ( 0.5) X 105 psi. Dental implants having spherical-pore diameters of 275 and 460 micrometers (/-cm) and cylindrical-pore diameters of 450 ,um were fabricated from this alloy. Dental implants were also fabricated from alumina porcelain ceramic composed of 95% aluminum oxide and 5% nepheline syenite. Material porosity is achieved by volatilizing void-forming cellulose particles within the ceramic.6 The resulting pore size and interconnectivity are not as controllable as with VMC, and pore diameters range from 200 to 400 ttm. The modulus of rupture under tensile loading of 75% dense material is 8,000 to 11,000 psi. Cylindrical implants of the two porous materials were surgically implanted into fresh and healed alveolar sites of extracted premolars in the mandibles of miniature swine. Typical implants, approximately 5 mm in diameter and 18 mm in length, are shown in Figure 1. Note the solid top of the implant that acts as a barrier to penetration of the porous matrix by oral fluid and debris, the removable screw that keeps a threaded channel in the implant clear during the
85

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

86

KARAGIANES ET AL

J Dent Res January-February 1976

FIG

12 13; f4 1.-Left, cylindrical-pore VMC dental imu-ight,

porous ceramic implants with posts.

plants; and

trau-sgingival

hone ingrowtlh period, and the transgingival post tllat is secured to the implant six weeks after implantation.

FIhe possible use of the nonocclusal transUIV gingival posts (also made of Ti-6A alloy) for attaclhment of artificial crowns is olvious; lhowever, their primarh y purpose in these experimenits was to stu(ly hone ingrowth reaction to nonocclusal intraoral stresses transmitte(d to the implants and to inivestigate gingival tissue reaction suirotunding the post. In contrast to thle slend(ler cylinclrical transgingival posts, several truLncated conical devices were also attachied to porous implants to investigate gingival response to this particular (lesign (Fig 2). In order to evaluate the effects of lhighler masticatory stresses on hone ingrowth, titanium alloy caps (Fig 2) were screwed on threaded transgingival posts and placedl in occlusion with the opposing maxillary teeth.
Free-standing gold crowns were attaclhed to two VMC dental implants to more accuirately assess the effects of normal occlusion on the bone-implant union; hlowever, these prostheses have not yet been fully evaluiated. Standardized and rapid techniques for oral

removable screw when the implant is tapped into place. Using the healed-site techniuLie, at least thiree montlhs were allowed for hone healing after tootlh extr actioni. The gingiva over eli(lentLetlous areas (contralateral mandibildar 1)1remolars 1, 2, and 3) was then incised and suifficiently elevated from the mandibular ci-est to allow space for surgical implantation. TIhe exposed hone was drilled using a standard orthopedic intramedullary pin (irill withl ai Jacobs-type chiuck and an appropriately sizedl twist (Irill hit. Using a hit 3 to
CAP SCREWDRIVER SLOT VENT 6

1
9.5 mm

7.6 mm ~~~~~~~~6.710O

TYP

implan-t surgery in swine were developedl in thiis program. Dental implant surgery was
followed in six weeks by transgingival post or cap attachment to the implant or both. This two-phase operation is believed necessary in swine becatUse of the difficulty of providing proper postoperative hygiene. Fess surgical instiuments are needed for the procedure and only one is specialized-a plutncll wiitl a recessed tip to prevent mutilating the

oo 000

ooo000000
000000

mm000 00 0 000
000

00 00 00 00 0 000 00 00 00 00 0 000

000000 00 0 000 000000 000000

000000

K4.8Bmm+TRUNCATED CONE

K4.8 mmADJUSTABLE OCCLUSAL CAP

Flm 2.-Diagraai of truncatedl cone and occlusal cap attachecd to VMC dental implants.

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Vol1 55 No. 1

POROUS IMPLANTS IN SWINE

87

5% smaller in diameter than the implant results in a tiglht interference fit l)etween the device and the bone. The site was liberally flushed with sterile physiological saline and the dental implant, autoclaved at 135 C for 35 minutes at 15 psig in triply distilled water, was then tapped snugly into the cavity. The gingiva was sutured over the implant, the top of whiclhI lies 1 to 2 mm l)elow the mandibular crest. In the fresh-site approach, implants were immediately placed in the freslh sites of ex tracted )remolar teetlh. Some alveolar bone removal was necessary in order to accommodate the cylindrical implant. All surgical and postoperative procedures were oteierwise identical to those used witlh the lhealed-site method. Th-e swine were kept indivi(lually in spaciouts, solid-walled pens to prevent selfmutilation- of the implanit sites. Thley were fed a soft diet for two weeks to allow gingival healing, and thien were retut nie(d to a standardl hard-pellet swine diet. The swine were anesthietizedl to attaclh transgingival posts or caps or both to the implants six weeks after initial implant smrgery. Tihis r apidl procetlutre coinis'ted of incising and elevating tlhe gingiva fItom the imlplant site an(d burring away lone that lhald growni over the tops of the implants. Thlc small implant screw was reThoved and re-

placed by the transgingival post, around whiclh the gingiva was approximated and sutured. No supplemental bracing was used to stabilize these devices. The six-week time intersal before attaclling posts and caps allows tissue ingrowth and bone organization to occur in the porous implants, giving sufficient strengthi and immobilization for retention. The experimental implants were evaluated by clinical and histological examination and mechalnical testing. Clinical examination it-n cluidedl inspection, pei-iodontal probe measuirements of the gingixval sulci, application of force to determine implant mobility, and periodic radiography to examine mandibular b)one response to the implants. Histological examination was performed oni nondecalcified sections (implant plus bone) that were delhydlratedl in graded alcolhols, infiltrated withl a low-viscosity embedding medlitum as described by Spunr,7 and stained uLsing techniques reported by Smithi and Karagianes.8 Soft tissue ancl decalcifiecd bone specimens were fixed in Formalin, embedded in paraffin, andl stained with hiematoxylini and eosin (H&E). MeTliaciical testing was performed oni fresh sectionis to letermine the slear strengthi of the hone-impllant interface. Thlie specimens were produced by cutting transverse sections

FIG 3.-Free-standing gold crowvn (attached to VMC implant) in full occlusion with maxillary teeth. Large distal embrasure promotes self-cleansing.
Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

88

KARAGIA4NES ET AL

J Dent Res January-February 1976

(4 to 9 mm thick) from the central region of the bone blocks containing the implants. The "pluLg" of implant material was then pressed from the surrounding bone using a tension-compression testing maclhinea antI the peak loads recorded. Interface shear strengths were calculated from specimen climensions andl peak loads. A total of 76 poious VMAC and ces-amic dlental implants were suLrgically implanted in the freslh andl healed sites of extracted mandibular pl)emolar teethl in 16 miniatur e swine. Twenty-three VMC and 12 ceramic implants were implanted in freshi premolar extractioni sites; 23 VlMC and 18 ceramic implants were implanted in lhealed extraction sites. Th-e swine were killed at appioximately 4, 5, 8, andl 12 monthis after implantation. Results To date, 68 dental implants have been examined. Nineteen were removed and examined at approximately four montlis, 17 at five monthis, 16 at eiglht monthis, and 19 at 12 months after surgical implantation. Four (onie VMC and three ceramic) were removed before the swine were killed because of rejection that eventually resultedl in more than 1-mm implant mobility when stressed manually. Eight othler prostheses, incluiding two VAMC implants with gold crowns attaclhed, remain to be evaluated at 17 months' postimplantation (Fig 3). Histological examination of nonldecalcified sections embedded in plastic showed bone ingrowtlh into 39 of 41 VMC dental implants. One had a thin fibi-ous tissue attachment and anotlher, as noted previously, was removed before the swine was killed. Nearly complete lhard tisstue invasion had occurred in several of the porous metallic implants and all but three had good bone attachment and ingrowtlh; tlhese tlhree had bone ingrowth over less than 25%-O of the bone-implant interface and exhibited less than 500
No adverse celltular inflammatory response cotild be fouind around any portions of the devices having intimate bone-implant adherence (Fig 4). T issuLe invasion and organized calcificaton appeared to follow a pattern of intramembranous ossificationi. Clusteis of osteoblasts were found sturrounda Instron Universal Testing Machine, model FCL, Instron Corp, Los Alamitos, Calif.

ing bone trabeculae that penetrated the vosids of the implants. Haversian systems were present in the invading bone. I hiin (lecalcified sections stained withl H&E (after implant pushioult tests) confirmed that normal bone was also present arouind the implants. Histologically, no morplological differences in the bone were observed in either the spherical or cylindrical poi-ous VATC implants he tween nonistressed, sliglhtly stressedl (transgingivxal posts) anid lhighily str-essed (occlusal caps) implants. No apparent diffesenices were noted between hone ingrowtli in implants from freslh or healed -sites, andl extenisis e tissuLe invasion anti calcificationi hiad occurred in four-montll implants as well as in those stutdlied for the longer time periods. Histological and mechanical testing resu-lts weie similar for 2715- and 450-4m pore size V7IC implants. Thlree of 27 ceramic (aluminla porcelain)

psi sheas- strength.

Fic. 4.-Histological cross sections showing ex-

tensive bone ingrowth into (top) 460-gm spherical-porec VMC implanit after fise months implantation (x50) and (bottom) 450-,1m cylinduiical-pose VIMC implanit eight months aftel surgery (Alizarini red S and methylene blue stain;
x20). Metal is black.

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Vol 55 No. I

POROUS IMPLANTS IN SWINE

89

1 :.

iA-" I

Fi. 5.-Histological cross sectioni of alumina poicelain implant showing bone-ceramic interfacc after five months implantationi. Lack of initercoinnectitig poiosity in material is exident (Alizarin red S and methylene blue stain; X20) .

implants failed and were removedl before the swinie were killedl. Histological examination was performed on nonidecalcified sectionis of
24 ce iamic implants removed at the (liffereiit

weie (luite firmly immobilized (less than 0.5 mm in)ovement) in the manidible and lacked any inif-lammatory responise indicative of im

stLudy intervals. Seven hlad an excellent boneimplatnt interface uinion (Fig 5), also suhstantiated byIi higlh mechanical shear sts-ength tests. Anotlher implant, not mechanically tested, also showed good bone invasion inlto implants was similar to that tthe VNIlC implants; hiowever, smaller amouLnt of lhardl tissue infiltration as a restult of the lack of adeqlutate interconnecting porosity in this material. Againi, n-o apparent differences in bone morplhology were noted in freslh vs lhealed sites, time periods investigated, or varying stress conditions placecd on these implants. The i-emaining 16 ceramic implants slhowedl zero to minimal bone ingrowthi and were characterized by a soft tissue interface union. Tlils fibrouis tissue rangedl from 100 to 500 Mm in tllickness and in most instances entirely sui roundled the implant. These hlistological observations were fturther substantiatedl 1y the low shear strengtlhs (O to 300 psi) of 13 of these implants. It slhouild be noted, hlowever, that these implants, as well as those anclhored by bone ingrowth.
the pores of these eighit founid witlh thete was a
the ceramic. Bone

Mechanical (pusliout) tests of the po0ous rlental iml)lants shlowedI a widle range of shiear strengths. Samples from some swine slhowed high interface shear strength values equixvalent to a total load-bearing capacity
of ulp to 1,000 11). It was evident from histological examination that the samples exhibiting low slhear strenigth valuies eithler lhad fibrous tissuie suirroun(ling the implant section or only spai-se bone infiltration over the suirf ace of the implant. Specimens slhowinlg hiigh shear strengtlh values demonstrated the expected complete incorporation of the implant section into mandibular bone, with extensive organized bone ingrowthi and absence of fibrous tissue. Figure 6 shows the bone-imlj1ant interface strengths of sections of 50 of the porouis implants puslhed ou-t on the tension-compression testing machine. Twenty-seven of the 30 VMIC implants tested had hiigh shiear strengths (ranging from 500 to 2,500 psi) inldicative of good l)one-implant interface union. In contrast, strengthls of 1.3 of 20 porotus ceramic implants were less than 500 psi (most less than 100 psi) as a result of little or no bone in-

Pl)anit iejection.

reslonrse

to

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

90

KARAGIANES ET AL

J Dent Res January-February 1976 terfered with a true evaluation of the gingiva-metal interface response. The extent of gingival tissue recession and inflammationi appeared to correlate directly with the amount of calculus on thle intraoral extensions. Generally, transgingival posts and occlusal caps lhadl tolerable amounts of calculus, and the gingiva-metal interface reaction was acceptablle, showing moderate epitlhelial liyperplasia and inflammatory response in the immediate area (Fig 8). Plasma cells macle up the majority of the inflammatory cells in these sites; some neutrophils and lymplhocytes were also present. Truncated-cone posts typically accumulated large calctulus deposits up to 1.5 mm in thickness. This cone-shaped design also appeared to dlirect the calculus in an apical dix ection, causing extensive gingival inflammation that itesuilted in excessive alveolar bone resorption (up to 2 mm in depth) in six of ten implants witlh these devices. In contrast, radiographic examinations plus m1(easuirements after the swine were killedl shiowecd alveolar bone resorption aiouind the tops of the majority of otlher implants to range from near zero to 0.5 mm. In two of the ssine dturing a one-year period, 1 2 im)lants witlh cylindrical transginlgival posts shlowed only zei-o to ap)roximately 0.2 mm hone loss (Fig 9) Epithelial invagirnation around the devices wxas a prxolem only isitlh the trunticated cones; the mucosal epithlelium was foiund to extendl ah)ically to the tops of the implants. Those witlh slendclei cxlindrical posts an(d occlusal casl)5 hl(I periodontal probe measurements ranging from 1.5 to 3.0 mm. (Gingival sulci measurements taken at six points on a num-

13

Q CERAMIC

[19

2000 2500 1500 3000 1000 500 BONEIIMPLANT INTERFACE SHEAR STRENGTH, psi

FtcG 6.-Mechanical test data for SO implants sloio'ing unumber ill each 500-psi shiear stress
i}icremen t.

growtl. Tlhis listological resullts

of

were coim-

patilble witlh

these

findinigs,

as

the

cer amic im)lants

xsere stii-rouLd(led

majotity
x

soft

tissule s-lather than hone.

1'lhe attaclhment of ttransgingival plosts. capt)s (placing the implants uinder varying stress conlitionis) made it possible to investigate ginlgval teaction to the titatiitm alloy deevices. Howevex, the rapidl accumulation of calculus on many of these exten-sions (ft-om lack of postoperative liygiene), as seen- in Figure 7, intt uncatcd coIes, atndl occlusal

F1ic 7.-Intraosal attaclhmenits (triuncatedI conie, occlusal cap, antI traizsginigival post) eight aftei implanitatioin. Note heaxy calculus deposits (left) Aftei calculus removal, ohservedI weise gingival inflammiiiationi ant(I ecession around toxic, excellent tissute t-esponise to cap and(I mild reactioni to small post (right) .
iniouths
Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Vol 55 No. I

POROUS IMPLANTS IN SWINE

91

FIG 8.-Histological sectioII of ginigiva at (arrow) slhowing epitlhelial inxagiiiation

tissue-transgingiral post interface andl iniflammatory cell responlse

(H&E staini; X140).

hei of

pr1erniolar teetlh in control swine iranged fi-om 1.0 to 2.5 mm.) A healthy und (lerlying soft tissUe-metal adheience impeded epithllelial migrationi, at leatst utp to the time
periodls insvestigated.
Discussion

Peihaps the most signiificant aspects of tills reseatrclh ai-e the encout aging resuilts telating to the possihility of long-term r etenitioii of enclosteal implants by an intimate hone-implant un1ioIn. hriere was no appartent rliffeietice in hone teactionls wlietlei- the im-

llants wsiee untler slighit stress or uniicler the liglh stiesses of occlusal contact. Ilie intent of mechanical testitig was to (leteimiite the hone-implant inteitface shieatstrengtlh and compat e and correlate these dlata, wlitlh histological observatioItis. The ilevasace of liglh shlear strengtlh values and tlteit relationshlip to load-bearing capalili ty and long-term retention of hone-ingr own iml)lants lhave not heen demonstrated; liowexet, it is significant that implnts iunlet hiighi stress (occlusal contact) slhowedl inter face strengths similar to tlhose uLndlet- slighit

FiG 9.-Radiographs showin-g (left) three

surgery.

VMC

implanits immedliately after transgingixval

post

Radiolucenit ate-as around tops of implaiits aie due to bon-e removal for post attachimenit. Righit, just hefoie (leath, show-ing miiiimal ahecolar hone resorption aftei one-sear implantation. (Dimensioiial distortions are caused by anigle of radiography.)
Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

92

KARAGIANES ET AL

J Dent Res January-February 1976

stress. For example, four VMC implants with occlusal caps, two at five and two at eight months after implantation, lhad shlear strengtlhs of 760, 1,700, 1,900, and 1,700 psi, respectively. These results, in conjunction witlh the histological findings, suggest that bone ingrowth is conducive to endosteal implant retention under masticatory con-

erative dental hygiene of the pig. Comparison of control swine gingivae with those at the implant sites showed that the inflammatory response of the latter is merely an exaggeration of that seen in the "normal" pig gingiva and couldl be classified, in most instances, as a moderate gingivitis.

ditions. The initial stability of porous devices implanted in bone is critical; however, it is not known what degree of movement prohiibits osteogenesis and initiates soft tissue encapsulation. Bracing devices were not used in thiese experiments, therefore initial stability was achieved at the time of surgery tlhrough a tight bone-implant interference fit. The hiiglh strength of the VMC material permitted this type of surgical fit and, with the high porosity of the material, resulted in organized bone ingrowth that securely anchored the implants in the mandible. The failure to obtain similar results with porous altumina porcelain implants appeared to be related to material porosity and strength. Increasing ceramic porosity (for improved bone ingrowth) resulted in decreased material strengthi and implant fracture when an interference fit was attempted. If implanted more loosely, initial stability was reduced and soft tissue encapsulation without bone invasion resulted. One might assume that bone ingrowth was initially present in some of the ceramic implants, with bone resorption occurring after the attachment of stress-producing intraoral devices; however, as bone was present in all but one VMC implant, the deduction was that hard tissue ingrowth did not occur at all in these ceramic implants. This is further supported by the fact that eight implants (three VMC and five ceramic with transgingival posts) in two of the swine during a one-year period had zero mobility but different pushout results. The three VMC prostheses had shear strengths of 2,300, 1,400, and 2,300 psi, respectively, whereas the shear strengths of all ceramic implants were less than 300 psi. Histological examination showed a soft tissue attachment to the ceramic implants in contrast to the bony attachment to VMC. Gingival reaction to the intraoral devices was not great, except with the truncatedcone design, and it was thought that much of the reaction reflected the poor postop-

Conclusions Organized bone ingrowth in endosteal porous implants fabricated from VMC titanium alloy and surgically implanted with a tight interference fit, securely anchored the implants in fresh and healed mandibular premolar sites of miniature swine. This boneimplant union retained its integrity under high as well as slight masticatory stresses up to one-year after implantation. Bone invasion of the alumina porcelain implants was impeded by the lack of adequate interconnecting porosity; when the porosity was increased, insufficient ceramic strength prohibited a tight initial bone-implant fit. As a consequence, inadequate initial implant stability resulted in a soft tissue encapsulation of the majority of the ceramic implants. Histological examination and mechanical testing results were similar for bone-ingrown implants exposed to different experimental stresses for 4, 5, 8, and 12 months. Bone ingrowth and interface shear strengths were also similar in the different VMC pore sizes and shapes investigated. The design of intraoral attachments appeared critical, at least in swine where no

postoperative treatment was administered. Gingival inflammation and alveolar bone resorption caused by calculus were severe around truncated cone-shaped devices. Slender transgingival posts, occlusal caps, and crown restorations were less susceptible to calculus accumulation, resulting in a more satisfactory gingival and subgingival response. Excessive epithelial invagination was a problem only in implants with transgingival truncated cones. Good adherence of soft tissue to metal under the gingival mucosa prevented epithelial migration around implants with other transgingival devices. Alveolar bone resorption around the tops of bone-ingrown implants was minimal at the time intervals examined (up to one year); however, a definite conclusion should be delayed until longer-term implants under full occlusion are evaluated.

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Vol 55 No. 1
References NATIELLA, J.R.; ARMITAGE, J.E.; GREENE, G.W., JR.; and MEENAGHAN, M.A.: Current Evaluation of Dental Implants, JADA 84: 1358-1372, 1972. GALANTE, J.; ROSTOKER, W.; LUECK, R.; and RAY, R.D.: Sintered Fiber Metal Composites as a Basis for Attachment of Implants to Bone, J Bone Joint Surg 53A: 101-114, 1971. HULBERT, S. F; YOUNG, F.A.; MATTHEWS, R.S.; KLAWITTER, J.J.; TALBERT, C.D.; and STELLING, F.H.: Potential of Ceramic Materials as Permanently Implantable Skeletal Prostheses, J Biomed Mater Res 4: 433-456, 1970. KARAGIANES, M.T.: Porous Metals as a Hard Tissue Substitute: I. Biomedical Aspects, Biomater Med Devices Artif Organs 1: 171181, 1973. WHEELER, K.R.; MARSHALL, R.P., and SUMP,

POROUS IMPLANTS IN SWINE

93

1.

2.

3.

4.

5.

K.R.: Porous Metals as a Hard Tissue Substitute: II. Porous Metal Properties, Biomater Med Devices Artif Organs 1: 337-348, 1973. 6. MARSHALL, R.P.; KARAGIANES, M.T.; RASMUSSEN, J.J.; and WESTERMAN, R.E.: Development and Evaluation of Artificial Dental Anchors of Non-Natural Design Implanted in Miniature Swine, in First and Second Year Report to National Institute of Dental Research, June, 1971 to June, 1973, Contract No. 211B-00402, Richlarid, Wash: Battelle, Pacific Northwest Laboratories. 7. SPURR, A.R.: A Low-Viscosity Epoxy Resin Embedding Medium for Electron Microscopy, J Ultrastruct Res 26: 31-43, 1969. 8. SMITH, L.G., and KARAGIANES, M.T.: Histological Preparation of Bone to Study Ingrowth into Implanted Materials, Calif Tissue Res 14: 333-337, 1974.

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010