biomaterials for facial bone augmentation: comparative studies

Biomaterials for facial bone augmentation: Comparative studies

June Wilson and G . E. Merwin Bioglass Research Center, 1. Hillis Miller Health Center, University of Florida, Gainesville, Florida 32610

Presently no material is available which is entirely satisfactory for facial bone augmentation. These studies examine several of those already in clinical use, made from various polymers in solid, porous, and woven forms. Homograft bone has also been studied, as an implant material. All materials were used in situations for which they are currently recommended clinically. Bioglass (Bioglass is a trademark of the Uni- versity of Florida) implants, which are sug- gested for clinical USE, have been studied in the same model and results show that their surface activity provides a more satisfactory immobilization, both in the short and long term, than does the tissue ingrowth

on which most of the other materials depend. Results show that in this model as well as in clinical practice, porous and woven materials provoke in tissues a continuing cellular response which will always compromise long-term clinical success. Autograft bone has associated morbidity and is unpredictable with respect to its in- corporation into host tissue and persist- ance at the site. Bioglass, however, was immobilized successfully at both hard and soft tissue interfaces without the need for porosity, could be satisfactorily shaped in the operating room, and, in addition, had the bonelike hardness which is not provided by any other available material.

INTRODUCTION

A wide range of materials has been proposed and is being used by plastic and reconstructive surgeons in postsurgical, post-traumatic, and cosmetic applications.lI2 So-called ideal materials have been described but none exists. Au togenous bone and cartilage have been successfully used but their harvesting requires a second operation which prolongs the operating time and increases morbidity and risk.3 Homograft materials must be expensively processed, stored, and monitored to ensure materials which are as sterile, disease free, immunologically inert, and as safe as po~sible,~ especially since the advent of AIDS.

Alloplastic materials are widely available and various modes of attachment to over- and underlying tissues have been promoted. All of these materials share a major problem, that of long-term stability and continuing function. Neither fibrous encapsulation nor porous ingrowth has provided entirely acceptable long-term reliability.’ Mobility at the tissueiimplant interface is unavoidable where mechanical retention by ingrown tissue is the only stabilizing influence. Such movement can produce inflammation and the porous structure needed to allow ingrowth can compromise both the hard-

]. Biomed. Mater. Res.: Applied Biomaterials, Vol. 22, No. A2, 159-177 (1988) 0 1988 John Wiley & Sons, Inc. CCC 0021-9304/88/A20159-19$04.00

160 WILSON AND MERWIN

ness and the integrity of the implant. A particular requirement of a biomaterial for facial bone augmentation is that, to be successful, it must have a bone-like hardness, be amenable to contouring at the time of surgery, yet be immobilized at both hard and soft tissue interfaces. Studies in which Bioglass implants were used for facial bone augmentation have been re- p ~ r t e d ~ , ~ and preliminary results show that the controlled surface activity associated with these materials can provide the bonelike hardness and stability needed. Studies here reported examine a wide range of materials including autogenous bone and others for which clinical data exist, and compare them, in a dog model, with Bioglass implants.

EXPERIMENTAL DESIGN

Four sites were selected and 30 mixed-breed adult dogs weighing 15-20 kg used. Sites were nasal dorsum, right and left maxillae, and chin (Fig. 1) . The animals and sites provided a model which approximates, in size of implant and location, human clinical requirements. All materials were made as rough, disc-shaped, 1-2-cm-diameter blanks and contoured at surgery as required. Bioglass implants were placed on nasal dorsum, chin, and right maxilla after removal of the periosteum and on the left maxilla without removal of periosteum. All other implants were placed after removal of periosteum.

Bioglass implants were examined after 4, 13, and 26 weeks, all others after 13 weeks. Distribution of implants is shown in Table I. This distribution reflects two considerations deemed important when the experiments were planned. The authors believe that an overriding constraint in experiments using animals is that no animal be used to provide data which are predict- able.* For this reason experiments to prove the suitability of Bioglass were done initially in a group of animals in all four sites. The 18 dogs in that study each had four Bioglass implants. Only after this was demonstrated successfully was the series of experiments using commercially available materials and autograft bone begun. The remaining 12 animals each carried more than one material. However, the commercially available materials that were used are recommended for and used clinically in specific sites and for this reason were not implanted in sites where clinical experience shows a high proba- bility of failure. We believe that clinical experience with these materials allowed us to reduce the number of animals in these groups to a minimum for the purposes of these comparisons.

Materials

Solid materials

45S5 Bioglass is a transparent bioactive material with proven ability to Implants were blanks made at the University of Florida and bond to

BIOMATERIALS FOR FACIAL BONE AUGMENTATION 161

Figure 1. X-ray showing implants in place on nose, chin, and maxilla.

sterilized by ethylene oxide. They were contoured, sometimes extensively, by the surgeon at implantation using diamond burs and water irrigation.

Silastic* is nonporous medical grade silicone rubber. Implants were carved from a block using scalpels under sterile conditions, taking care to minimize

"Silastic is a trademark of the Dow-Corning Corporation.


sharp edges. They were then cleaned in hot water and soap solution to remove surface contaminants, rinsed in sterile water, wrapped individually in lint-free material, and sterilized by autoclaving.

Autogenous bone was harvested acutely from the right chest under sterile conditions. Approximately 5 cm of rib bone was harvested and its periosteum removed. Segments 2.5 cm long (for nasal dorsum implants) and 1.5 cm long (for maxillar implants) were cut and placed in sterile saline until needed.

Porous ma ter ia 1 s

Medpor** is ultra-high-density polyethylene with large pore size of 100-500 pm the pore size being large enough to allow ingrowth of bone rather than soft tissues. It was supplied as a presterilized block 38 x 63 x 6.5 mm with a prohibition on resterilization. Implants were therefore carved under sterile conditions at the time of surgery.

Nylarnid' is a fabric of polyamide fibers made from twisted fibrils which are then woven into a fabric. A 30-cm square of fabric was sterilized by autoclaving. At surgery, under sterile conditions, strips 2.5 cm X 2.5 cm were cut, folded into thirds, and then rolled to form the implants. The free edge was sutured with 4-0 polyamid sutures as described by Stucker."

Proplast * is a composite felt material consisting of polytetrafluoroethylene reinforced by pyrolysed carbon. It has 70-90% pore volume with pore sizes from 80-400 pm. The material is black in color. Implants were made from a block of Proplast by carving under sterile conditions taking care to avoid crushing and consequent alteration of the pore structure. After rinsing in sterile water they were sterilized by autoclaving.

Methods

Surgical

Mixed-breed dogs of 15-20 kg were given an intramuscular injection of ampicillin (15 mg/kg) before anesthesia with Biothal induction and endo- tracheal administration of halothane. Operative sites were shaved and swabbed with Betadine (10% povidone iodine) and draped. Rib was harvested first and the chest closed.

Chin sites were exposed, periosteum removed, and implants placed in accordance with manufacturers' instructions, Proplast being first soaked in 10% penicillin solution. Skin was closed with 3-0 silk sutures.

Nasaf sites were similarly prepared and a subperiosteal pocket made through an incision at the distal end of the bony dorsum. Bioglass implants,

*"Medpor is a trademark of Portex Inc. 'Nylamid is a trademark of S Jackson, Inc. *Proplast is a trademark of Vitek, Inc.


cast as blocks 2 X 1 cm, were shaped further so that a projection on the long axis was formed which would slide into the suture between the nasal bones. Rib grafts approximately 2.5 cm long were shaped to fit. Nylamid implants were soaked in Lincocin (Lincomycin hydrochloride 300 mg/ml) ends shaped with scissors to fit and placed. Skin was closed with 3-0 silk.

Max-illary sites were exposed and in most cases the periosteum elevated and removed. All implants were placed on the bone, slight modifications in size and shape being made as necessary. Proplast and Medpor were first soaked in pcncillin solution. In 18 animals one maxillary site was exposed without removal of periosteum and Bioglass implants placed on the periosteal surface. Skin was closed with 3-0 silk. None of the implants was spe- cifically immobilized at surgery and no precautions taken postoperatively to protect the implant areas. Animals were kept in holding areas for 2 days postoperatively on a soft diet and given four doses of Ampicillin (15 mg/kg) intramuscularly at 6-h intervals. They were returned to individual pens after 48 h.

Animals were examined regularly for signs of infection or reaction at rib harvest or implant sites. Sites were x-rayed immediately and at 4-week intervals using standardized techniques.

Autopsy

Animals were killed by an overdose of anaesthetic and exanguinated. Sites and surrounding tissues were examined and gently manipulated to deter- mine mobility. Assessment of the type of attachment, where detected, was made. Implants and surrounding tissues were fixed in formalin and glu- teraldehyde fixatives as appropriate. Sites which contained Bioglass were later bisected and half of each sample decalcified before embedding in paraf- fin and sectioning. Remaining halves were embedded in gIycolmethacrylate and sections cut using a Leitz 1600 sawing microtome. Sections were stained with haematoxylin and eosin, Masson’s trichrome, and others as required.

RESULTS

Clinical

One animal with Bioglass implants developed an infection shown by culturing to be Staphylococcus mreus, which caused the loss of three implants, extruded from the infected sites. The nasal implant remained stable.

Five further Bioglass chin implants were lost by extrusion during the experimental period, however, in these cases there was no evidence of infection. One of six Proplast implants and two of six Silastic implants in the chin developed purulent discharges. After culture with sensitivity plates, treatment with lincomycin, by mouth, 22 mg/kg every 12 h for 7 days resolved the condition and the implants were retained.


Two Bioglass maxillary implants were not found at autopsy. Three animals developed seromas at the rib harvest site which resolved with needle aspiration.

Radiographs showed no evidence of underlying bone resorption in any case.

Macroscopic

At autopsy all sites were examined for signs of inflammation, infection, and other changes. Scarring was noted in association with chin implants of Silastic and Proplast. Some local discoloration was associated with Proplast. All implants were palpated in situ to assess their stability. Of the materials studied only two, Bioglass and autogenous rib are capable of establishing a physicochemical bond with hard tissues,i2 and an interface which does not allow micromotion. All the others can be stabilixd mechanically by ingrowth of soft tissue, where pores are <100 pm, or by bone into pores of greater dimension^.'^ In the larger series of implantations with 45% Bioglass no differences were seen with respect to bonding between 1-, 3-, and 6-month groups. When bonding takes place it does so in the first month, data on 1, 3, 6 groups were therefore combined.

Stability data overall are seen in Table 11. Since not all materials were applied to all sites the influence of the site on the achievement of a stable implant must be considered. In Table III the influence of situation becomes

TABLE I Distribution of Implant Materials

Nose I<. Maxilla I>. Maxilla Chin

Solid materials Bioglass 18 18 18" 18 Silastic 0 6 0 6 Bone 6 0 6 0

Porous materials Medpor 0 0 6 0 Nylamid 6 0 0 0 Proplast 0 6 0 6

"These were the only implants placed on top of the periosteum.

TABLE I1 Clinical Stability of Implants

In Bone In Soft Tissue Unstable Lost Group Size

Bioglass 40% 35% 11% 14L'c 72 Silastic 0 0 100% 0 12 Bone 33% 17% 8% 42% 12 Medpor 0 100% 0 0 6 Nylamid 0 33% 67%- 0 6 Proplast 0 42% 58% 0 12

BIOMATERIALS FOR FACIAL HONE AUGMENTATION 165

TABLE 111 Clinical Stability in Different Sites

Bone Soft Tissue Unstable Lost Group Size

a) Nasal Bioglass Bone Nylamid

b) R. maxilla Bioglass Silastic Proplast

c ) L. maxilla Bioglass" Bone Medpor

Bioglass Silastic Proplast

d) Chin

100% 66%

0

50% 0 0

11% 0 0

0 0 0

0 17% 33%

33% 0

83%'

720; 0

100%

33% 0 0

0 17% 67 B

6 0 17% 0

35°C 100% 100%

0 0 0

11% 0 0

11% 83 %

0

18 6 6

18 6 6

18 6 6

18 6 6

"Over periosteum

clear, successful results being more likely on the nose than the chin, at least in this model. The difficulty of preparing the surgical site without damaging the periosteum is evident in those implants on the maxilla which were placed on top of periosteum. However, observation on those implants placed after periosteum was removed showed a similar inconsistency.

Clinically stable implants were achieved with some consistency in all sites but the chin. Most lost specimens were from this site and those implants which were retained did not achieve stability. The loss of five Bioglass chin implants was discovered only when x-rayed after 4 weeks. A transgingeval approach, such as is used clinically, was used initially in this experiment, however it was replaced by a transdermal approach as soon as retention problems were identified. The loss of these implants was, we believe, due to the unsuitability of this particular surgical approach. Bone implants which are designated as lost were probably resorbed, since even those bone implants which survived were so reduced in size at retrieval that they were identifiable only after microscopic examination.

Microscopic

Bioglass

Implants were described clinically a s "bonded to bone" when the implant was immovable relative to the underlying bone. Histological examination did not always detect sites of direct bone bonding. The most usual response was regeneration of periosteum between bone and implant, adherent to both surfaces and providing a functioning connection between them which did not permit relative movement (Fig. 2). In many cases bone was present


Figure 2. Original periosteal tissue (A) between the nasal bones compared with regenerated tissue (B) between bone and Bioglass implant (C) (original magnification x 200).

directly at the interface and bonded to the implant surface (Fig. 3). Both types of response were seen along the surface of the implants.

Bonding to soft tissue was noted macroscopically when the implant was stable in and firmly attached to the surrounding soft tissue such that the capsule or covering needed forcible stripping to remove it. This was seen in


Figure 3. magnification x 200).

Bone at the interface (+) with a Bioglass implant (B.G.) (original

the soft tissues overlying the bone-bonded nasal implants as well as at other sites, When the surface of the implant, from which the adherent tissue had been stripped, was examined in the SEM, collagen fibers could be seen still attached to the Bioglass surface (Fig. 4). At the light microscope level fibroblasts at the interface were arranged perpendicular to the implant surface with fibers inserted into it (Fig. 5). Other samples and other locations


Figure 4. herent collagen (original magnification x 2,500).

Scanning electron micrograph of the implant surface beneath ad-

showed a more usual parallel arrangement of the collagen fibers at the surface but with the separation of fibers within the capsule which has been attributed to the effect of bonding of the inner layer to the implant14 (Fig. 6). The tissue response to unstable Bioglass implants was typical of that to an inert material, that is a capsule of varying thickness and cellularity, the



Fibroblasts perpendicular to Bioglass implant site (B.G.) (original

thickness and cellularity being greater in samples which had been particu- Iarily mobiIe during the experimental period.

Silastic

The tissue response to Silastic varied with the site, those on the maxilla were contained in a thin, relatively acellular capsule with no obvious fluid.


Figure 6 . (B.G.) (original magnification x 200).

Collagen fiber peeling from capsule around Bioglass implant

There was some osteoclast activity in the underlying bone in two cases. Although designated "unstable" because all were mobiIe to some extent, in one case the capsule was so thin that bone appeared to be present at the interface (Fig. 7). In the chin, however, the tissue response was indicative of continued trauma and delayed wound healing.

BIOMATERIALS FOIZ FACIAL BONE AUGMENTATION 171


Bone (B) at the interface with a Silastic implant site (S) (original

Bone

In four of the nasal implants focal attachment to the underlying bone was seen and in three cases revitalization of the graft had begun. Most of the graft material at 3 months was nonviable. MaxiIIary implants had almost all disap-


Figure 8. polarised to demonstrate the polymer) (original magnification x200).

Giant cell (-1 at the interface of a Medpor implant (M) (slightly

peared, a small fragment of one persisted but had not revitalized and w’as unconnected to the underlying bone.

Medpor

Although all six implants were considered stable in soft tissues, two were slightly more mobile than the others. No distinction was possible, micro-

BIOMATERIALS FOR FACIAL BONE AL'GMENTA'IION 173

scopically, between them. All implants were filled with relatively acellular fibrous tissue. Some giant cells were present throughout the implant (Fig, 8). In spite of the size of the pores (100--500 pm) no bone was seen in any implant in this experiment.

Nylamid

Microscopically the fabric mesh was invaded throughout by tissue of variable cellularity which tended to be more collagenous between the threads and cellular within the threads.

The cellularity was largely phagocytic with numerous giant cells and macrophages. Between the threads the connective tissue was often well vascularized but no capillary invasion into the threads was seen. There was no attachment to bone and no distinction possible between stable and unstable implants. Erosion of underlying bone was frequently seen.

Pvoyiast

In the maxilla, scarring was seen, especially over the mobile implants. Ingrowth of tissue was variable in both amount and cellularity, tending to contain predominantly fibrous tissue adjacent to the bone and to be very cellular on the subcutaneous side with giant cells, macrophages, and leuco- cytes present. There was also focal hemorrhage and necrotic tissue within the pores. No bone had grown into the implants and focal erosion of the underlying bone was seen in two of the six cases. Proplast in the chin was very unstable. Although there was well vascularized fibrous connective tissue at the periphery of the implant the bulk was filled with phagocytic cells, both macrophages and giant cells. Adjacent tissues contained granulomas associated with fragments of dispersed material (Fig. 9) and macroscopic discoloration of adjacent tissue was due to the carbon component which had become dispersed.

DISCUSSION

A variety of materials is available to surgeons for augmentation and reconstruction of bony tissues in the face and head. Autologous bone has provided a source where procedures are lifesaving and critical but harvesting of a rib is understandably not perceived as desirable by those who seek augmentation for aesthetic reasons. Materials presently available are predominantly porous or woven and depend on tissue ingrowth for stabilization. By virtue of their composition (usually polymer) and texture they do not have the bonelike hardness of an ideal biomaterial for this application.

No material performs ideally in all situations; those currently used are more suited to certain sites than others. Selection of the most appropriate material for a particular patient depends on the best information currently available to the surgeon. This information derives from published work,


Figure 9. Granuloma associated with dispersed Proplast: light slightly polar- ized to demonstrate polymer fragments in giant cells (-+) (original magnification ~ 2 0 0 ) .

scientific presentations, manufacturers’ promotional literature, and word-of- mouth from surgical colleagues. In this series of experiments, materials have been used in the sites for which they are currently favored and compared with Bioglass implants.

BIOMATENALS FOR FACIAL BONE AUGMENTATION 175

The concept of stability is one which we have considered in two ways. Clinical stability is the most desirable surgical endpoint, the biomaterial functions as well as the original bone, is immobile and enduring. However, as we have observed, clinical stability was not always supported by histological observation, a "stable" implant interface, even after several months, might display more cellularity than expected if the interface were a genuinely stable one. The presence of inflammatory cells including plasma cells, poly- morphonuclear cells, and various phagocytes suggests that long-term clinical stability is compromised. Histological assessment of the interface, whether deemed clinically stable or not, allows discrimination of reliable from less reliable materials.

The ideal implant material for bony augmentation and reconstruction should be as close as possible in form and function to the bone it mimics or replaces. After healing, the tissues surrounding the implant should interface with the implant as they would with natural bone. We have shown that for Bioglass implants this is the case. Those materials which depend on mechanical fixation invariably had a cellular response indicative of continuing micromovement at the interface. The most rigid materials, Silastic and Medpor, showed the most stable tissue response microscopically. Especially interesting was the demonstration that Silastic, if effectively immobilized, allows bone to grow to the interface, without a capsule which can be distinguished at the light microscope level.

The woven material Nylamid demonstrated the effect of micromovement on the newly formed connective tissue. Within the threads of the woven fabric the tissue was unstable and cellular, between the threads, where relative movement is less, there was a more stable tissue with a good blood

Proplast, a composite feltlike material, is very easily contoured but much more sensitive to unsatisfactory initial immobilization. A stable interface was seen only focally and fragmentation and dispersion of surface fragments and attendant tissue response was a consistant problem in the sites in which we used it.

Autologous bone did not, in these experiments, provide a useful material for maxillary augmentation, apparently> being resorbed under pressure. On the nose, it was more promising but was very slowly revitalized. Osteoclast activity under stress is a well known response of bone15 and it was seen in these experiments associated with widely differing materials, Silastic, Nylamid, and Proplast as well as the autologous bone implants.

"Osteoproduction" was seen associated with the Bioglass implants. The proper distinction of osteoinductive and osteoconductive materials con- tinues to be discussed.16 Bioglass is not an osteoinductive material, as count- less soft-tissue implants without bone production have shown, although it is osteoconductive. Additional bone, which is not continuous with underlying bone, appears at the Bioglass implant surface. This occurs in sites which have been drilled or contoured and within which cells with osteo- genic potential have been liberated from periosteum or Haversian systems.

supply.


These cells colonize the Bioglass surface on implantation and produce bone, thus speeding the eventual stabilizing of the implant in bone. This process, complementing osteoconduction, we have called ”osteoproduction.”

The superiority of chemical over mechanical attachment is confirmed in these experiments, although the need for primary immobilization during the initial bonding period is also demonstrated, the chin in this animal model proved a totally unsuitable site. Success may be possible in a patient who can help to protect the site in the immediate postoperative period. Dogs, however, cannot be persuaded to do this. Nasal implants which fitted into the nasal suture were most successfully immobilized and eventually satisfactory.

The ability of Bioglass to bond to both hard’ and soft16 tissues with elimi- nation of micromovement at the interface, its bone like hardness and its ability to be contoured’‘ in the operating room provide advantages over materials presently in use which make it a valuable addition to the range of options available to surgeons concerned with facial bone augmentation and reconstruction.

The authors are grateful to C. Maas, M.D., L. Rogers, M.D., M. D. Frey M.D., and R. Martin M.D. for their surgical skills and encouragement and to L. L. Hench, Ph.D. and D. Spilman, B.S. for supply of Bioglass implants.

Financial support for these studies was provided by American Biomaterials Cor- poration, Princeton, N. J.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

J. W. Kent and M. F. Zide, ”Wound healing, bone and biomaterials,” Ofolaryngol. Clin. North A m . , 17, 273-319 (1984). A. D. Kornblut, T. W. Stark, J. G. Vap, et al., ”The role of autografts, homografts, heterografts and alloplastic implants in reconstructive head and neck surgery,” Otolaryngol. Clin. North Am., 15, 147-160 (1982). J. J. Longacre, J. M. Converse, and D. M. Knize, in Transplantation of Bone in Reconstructive Plasfic Surgery: General Principles, 2nd ed., Vol. I, W. B. Saunders Co.,J. M. Converse (Ed.) Philadelphia, 1977, Chap. 13. R. C. Burwell, “The capacity of fresh and treated homografts of bone to evoke transplantation immunity,” J. Bone Joinf Surg., 45B, 688-710 (1962). 8. L. Brown, H. B. Neel, and E. B. Kern, “Implants of Supramid, Pro- Plast, Plastipore and Silastic,” Arch. Ofolaryngol., 105, 605-609 (1979). G. E. Merwin, L. W. Rodgers, June Wilson, and R. G. Martin, ”Facial bone augmentation using Bioglass in dogs,” Arch. Otolaryngol. Head b Neck Surg., 122, 280-284 (1986). J. Wilson, G. E. Merwin, L. Rodgers, R. Martin, and D. Spilman, “Facial bone augmentation using Bioglass in dogs,” in Biological and Bio- mechanical Performance of Biomaferials, Christel, Meunier, and Lee (Eds.), Elsevier Science Publishers Amsterdam, 1986, 93-98. L. L. Hench and E. Ethridge, Bioniaterials: A n interfacial Approach, Aca- demic Press, New York, 1983, p. 288. L.L. Hench, L.L. Splinter, T. K. Greenlee, and W.C. Allen, “Bonding mechanisms at the interface of ceramic prosthetic materials,” J. Biomed. Mater. Res., 2, Part 1, 117-141 (1971). L. L. Hench and E. C. Ethridge, Biomaterials: An lnferfacial Approach, Aca- demic Press, New York, 1983, 139-148.


11. F. J. Stucker, R. H. Hrokawa, and C. W. Pruet, " The autoalloplast: An alternative in facial implantation," Otolaryngol. Clin. North Am., 15, 161-168 (1982). L. L. Hench and H. C. Paschall, "Direct chemical bonding between bioactive glass ceramic materials and bone," J. Biomed. Muter. Res. Syrnp., 4,

C. A. Van Blitterswijk, et al., "Macropore tissue ingrowth: A quantitative and qualitative study on hydroxyapatite," Biornuterinls, 7, 137- 143 (1986). D. R. Nolletti, J . Wilson, G. E. Merwin, and L. L. Hench, "Mechanisms of soft tissue bonding to Bioglass," Trans. Soc. for Biomaterials, IX, 143 (1986). W. Bloom and D. W. Fawcett, A Textbook of Histology, 10th Edition, W. B. Saunders, Philadelphia, 1975, 257-262. Definitions in Biornaterials: Proceedings of a Consensus Conference of the European Society for Biomaterials, Chester, England, D. L. Williams (Ed.), Elsevier, Amsterdam, 1987, p. 71. J. Wilson, G.E. Merwin, and L . L . Hench, "Machining Bioglass implants," SAMPE J., 21, 6-8 (1985).

12.

25-42 (1973). 13.

14.

15.

16.

17.

Received September 27, 1987 Accepted March 31, 1988

biomaterials for facial bone augmentation: comparative studies

Documents