J Oral Maxillofac
Surg
48:10@3-1074,1990
Tissue Response to Composite Ceramic Hydroxyapa titel Demineralized Bone Implants GAIL Y. PETTIS, DDS, MMEDSC,* AND
JULIE
LEONARD
GLOWACKI,
B. KABAN,
DMD,
MD,t
PHD$
This study evaluated the tissue reactions to two materials: ceramic hydroxyapatite (CHA), and a composite material of demineralized bone powder (DBP) and CHA (ratio of 4:l) in a collagen vehicle. The materials were tested in a subcutaneous pocket, a mandibular onlay, and in a calvarial onlay model. Specimens were evaluated histologically at 7, 10, 14, and 21 days postimplantation. Ceramic hydroxyapatite, implanted subcutaneously, elicited a fibrous response with minimal inflammation, but did not induce bone formation. In specimens of subcutaneously implanted composite material, induced bone was evident in association with the DBP. In CHA onlay specimens, there was a small amount of reactive bone extending from the host bone into the implant. In composite onlays, bone filled the entire body of the implant. The results of this study indicate that CHA particles were not osteoinductive in heterotopic sites and that osteoconductive ingrowth was minimal in onlays. Bone was induced by DBP even when mixed with CHA particles and implanted in subcutaneous and intraosseous sites. It was concluded that composite implants may provide a means of combining the osteoinductive properties of DBP with the bulk and structural support of osteoconductive CHA particles.
The atrophic mandibular ridge provides one of the most difficult challenges in the field of reconstructive surgery. Conventional treatment involves augmentation with onlays of autogenous bone. These grafts, however, undergo progressive, uncontrolled resorption during healing.’ Donor site morbidity is an additional drawback to the use of fresh autogenous bone.2 These disadvantages have stimulated the development of alternative implant
materials, such as particulate ceramic calcium phosphates.3 In addition to eliminating donor site morbidity, these materials provide immediate firmness and are available in nonresorbable forms.4 A disadvantage to the use of particulate ceramics is their tendency to migrate through the surrounding connective tissue.3 Furthermore, the nature of the tissue response to these materials is inadequately understood. Particulate ceramics provide a suitable substrate for bone deposition, a property known as osteoconduction. Controversy exists, however, regarding whether osteoconduction reaches the height of the ridge augmented with ceramics. Demineralized bone has been demonstrated as an alternative material for several types of osseous reconstruction.5-7 The mechanism of healing with demineralized bone matrix is osteoinduction. The matrix stimulates phenotypic conversion of connective tissue cells into chondroblasts with endochondral bone formation.839 Therefore, extraskeletal implantation sites may be used to distinguish between inductive and conductive mechanisms of healing.”
* Assistant Professor, Department of Orthodontics, College of Dentistry, University of Tennessee, Memphis, TN. t Professor and Chairman, Department of Oral and Maxillofacial Surgery, University of California School of Dentistry, San Francisco, CA. $ Associate Professor, Department of Orthopedic Surgery, Harvard Medical School, Brigham and Women’s Hospital, Boston, MA. Address correspondence and reprint requests to Dr Pettis: Department of Orthodontics, University of Tennessee, College of Dentistry, 875 Union Ave, Memphis, TN 38163. 01990 geons
American
0278-2391/90/481
Association
of Oral and Maxillofacial
Sur-
O-001 0$3.00/O
1068
1069
PETTIS ET AL
Kaban and Glowacki’ developed a rat mandibular model to evaluate demineralized bone matrix and other bone-derived graft materials for use in mandibular ridge augmentation. Various materials were placed supraperiosteally in the edentulous area located between the incisor and first molar. Unlike mineral-containing bone particles, demineralized bone matrix induced a mass of bone that was firmly united to the ridge by the 14th day after implantation. However, the practical drawback to the clinical use of demineralized bone implants is that, unlike ceramics, they do not provide immediate rigidity. In this study, we evaluated tissue reactions to particulate CHA and to a composite of demineralized bone powder and CHA in three sites. Tissue responses to implants of DBP have been well documented in animal models8.9*“; these sites were also implanted with CHA alone to determine the portion of the healing response attributable to the ceramic. Subcutaneous pouches were used for positive identification of an osteoinductive process. Osteoconduction was observed in mandibular onlays. The calvarial onlay site provided the additional benefit of direct access for complete removal of the periosteum. Methods and Materials
PREPARATIONOF IMPLANT MATERIALS Particulate CHA (Calcitite 2040) was obtained from Calcitek, Inc (San Diego, CA). The particles were dense spheres with diameters ranging from 425 to 850 km. Approximately 0.2 cm2 was used for each implant. Demineralized bone powder was prepared for composite implants according to the protocol of Glowacki and Mulliken.” Briefly, diaphyses of long bones were obtained from Charles River rats. After the bones were scraped free of soft tissue and marrow and thoroughly washed in cold, deionized water, they were extracted with ethanol followed by anhydrous ethyl ether. The resulting dehydrated material was pulverized in a liquid nitrogen impacting mill (Spex, Metuchen, NJ). Pulverized bone particles were sieved into fractions according to particle size. Particles in the range of 75 to 250 pm were demineralized with 0.5 mol/L hydrochloric acid (50 mL per g bone) for 3 hours at room temperature. Acid and liberated minerals were washed away with cold, deionized water (500 mL per g) until the pH of the wash matched the pH of the water. After extensive washings with water, the DBP was extracted with absolute ethanol followed by anhydrous ether. The DBP was mixed with CHA in a 4: 1 ratio by volume. The mixture (0.2 mL) was moistened with
approximately 0.1 mL of Zyderm collagen (Collagen Corp, Palo Alto, CA). SURGICAL PROCEDURES Both CHA and the composite mixtures were implanted in subcutaneous pockets and used as onlays for the mandible and calvarium. Sham operations were performed in onlay sites so that histologic changes due to surgical trauma could be evaluated. In addition, mandibles and calvariae were obtained from intact, age-matched animals for use as controls. All specimens were obtained at 7, 10, 14, and 21 days postoperatively. Site I: Subcutaneous
Pouches
Twelve 28-day-old male Charles River rats were anesthetized by inhalation of methoxyflurane. After the surgical area was shaved and swabbed with alcohol, bilateral ventral incisions were made on either side of the midline at the level of the diaphragm. Narrow cephalad pockets were created by blunt dissection and approximately 0.2 cm2 of implant material was deposited at the end of the pocket, 1 cm away from the incision. The incision was closed with a 9-mm wound clip. Site 2: Supraperiosteal Mandibular Onlays Twenty-four 28-day-old male Charles River rats were sedated and anesthetized by inhalation of methoxyflurane and intraperitoneal injections of Brevital (5 mg/kg; Eli Lilly, Indianapolis, IN). Lidocaine (2% with epinephrine l:lOO,OOO)was infiltrated into the submental region bilaterally. After the surgical area was shaved and swabbed with alcohol, a horizontal incision was made in the submental triangle just posterior to the mandibular symphysis. Both horizontal rami were exposed and a submucosal pocket created anteroposteriorly on the superior aspect of the edentulous region between incisor and molars, leaving the periosteum attached to bone. Test materials were implanted and deep closure achieved with 4-O chromic catgut suture. The skin incision was closed with two 9-mm wound clips. These animals received a diet of milled standard rat chow for 2 weeks following surgery. Specimens were harvested by excision through the body of the mandible in the molar region. Sham operations were performed on four rats. Anesthesia and soft-tissue dissection were performed as above; however, no implants were placed. These animals received a diet of milled standard rat chow for 2 weeks.
1070
TISSUE RESPONSETO CHA/DBPIMPLANTS
Site 3: Calvarial Onlays
Twenty-four 28-day-old male Charles River rats were sedated and anesthetized by inhalation of methoxyflurane and intraperitoneal injections of Brevital(5 mg/kg). After shaving and swabbing with alcohol, lidocaine (2% with epinephrine 1: 100,000) was infiltrated into the surgical area. A coronal incision was made at the level of the ears. The scalp was reflected anteriorly and periosteum removed from the bone. Implant material was placed over the sagittal suture between the frontal and occipital sutures. Incisions were closed with interrupted 4-O nylon suture. Sham operations included anesthesia and soft-tissue dissection, and complete removal of the periosteum. These procedures were performed on four rats. HISTOLOGIC METHODS After they were killed, all specimens were immediately fixed in cold 2% paraformaldehyde in 0.1 mol/L cacodylate buffer (pH 7.4) for 24 hours. Mandibular and calvarial specimens were demineralized at 4°C with frequent changes of the demineralizing fluid, which was either 7.5% EDTA mixed in 0.1 mol/L cacodylate or a solution consisting of 0.05 mol/L of formic acid and 0.03 mol/L sodium citrate. After demineralization for 3 weeks, specimens were rinsed thoroughly with cold, deionized water and embedded in glycol methacrylate embedding media. Three-micrometer histologic sections were made as follows, with a Jung Model 1140 Autocut microtome (Reichert-Jung, Vienna, Austria) fitted with a Carborundum knife. 1. Subcutaneous implants: The section was taken at approximately one-third to one-half the depth of the implant. 2. Mandibular implants: The section was taken in the coronal plane, at the anterior-posterior midpoint of the implant. 3. Calvarial implants: Each calvarial specimen was divided into two portions along the sagittal plane. Coronal sections were made approximately 1 mm from the cut edge of each half. Sections were stained with toluidine blue. Additional sections were stained for tarn-ate-resistant acid phosphatase activity (a marker of osteoclast activity) and for alkaline phosphatase activity, which served as a marker for osteoblast activity.12 The following features were quantitated in the sections: 1) the total area of the implant, 2) the area occupied by ceramic particles, and 3) the area of new bone within or adjacent to the implant. A mag-
nification conversion units of mm*.
factor was applied to obtain Results
SUBCUTANEOUS IMPLANTS
CHA. By day 7, each particle was enclosed by layers of fibrous connective tissue (Fig 1). With the exception of an occasional multinucleated cell, neither acute nor chronic inflammatory cells were evident at any time point. This indicated that acute inflammation that may have been associated with the surgical procedure or with CHA was resolved by day 7 and that CHA did not elicit a foreign-body reaction. No cartilage or bone was observed in any of the specimens. Staining for tartrate-resistant acid phosphatase and alkaline phosphatase activities was negative. Composite. Figure 2 demonstrates the distribution of CHA and DBP particles in the preimplanted composite material. Subcutaneous composite implant specimens were highly cellular with fibrous connective tissue surrounding each particle at day 7. By day 14, islands of induced cartilage were present between the DBP particles, but not in close proximity to CHA (Fig 3). By day 21, rows of osteoblasts and occasional foci of osteoclast-mediated bone resorption could be identified (Fig 4). Alkaline phosphatase activity was associated with the osteoblasts, and positive staining for tartrate-resistant acid phosphatase was identified in the osteoclasts, especially at their surfaces adjacent to bone. MANDIBULAR SPECIMENS
Control. Coronal sections at the anterior-posterior midpoint of the mandibular implantation site
FIGURE 1. Subcutaneous CHA implant, day 7. “CHA” denotes region previously occupied by ceramic particles. (Particles of ceramic were lost in the sectioning procedure. Ceramic remnants remain in these spaces.) (Toluidine blue, original magnification x 50.)
PETTIS ET AL
FIGURE 2. Preimplanted composite material. “CHA” denotes spaces previously occupied by the ceramic implant. DBPs are distributed throughout. with CHA interspersed ttoluidine blue, original magnification x38).
demonstrated the normal anatomy of the rat mandible. The incisor was located in the center of histologic cross sections through the control mandibles. The outermost layer of the alveolus consisted of lamellar bone with a regular circumferential pattern of organization. This bone was surrounded by thin, fibrous periosteum. Tartrate-resistant acid phosphatase and alkaline phosphatase activity indicated remodeling at the surface of this bone. Immediately superior to the periosteum was a thin layer of muscle covered by gingival mucosa and epithehum. Sham. The histologic appearance of the incisor and its supporting structures in the sham-operated specimens was identical to that of the control specimens with the exception of an occasional area of tissue separation between periosteum and muscle or gingival mucosa. These results indicate that any acute inflammation that may have occurred due to
FIGURE 3. Subcutaneously implanted composite material, day 14. Spaces previously occupied by CHA are located at the periphery of the field, with DBP particles in the center. Arrows denote areas of chondrogenesis located between DBP particles (toluidine blue, original magnification x38).
1071
FIGURE 4. Subcutaneously implanted composite material, day 21. CHA particle spaces can be seen at the periphery of the field. Arrows denote area of induced bone. (toluidine blue, original magnification x38).
surgical manipulation had resolved by day 7. No reactive woven bone was present. CHA onlays. The histologic appearance of the incisor and the bone surrounding it was similar to that of control specimens. However, at all time points, there were areas of newly formed bone continuous with the outermost layer of alveolar bone defined by the control specimens (Fig 5). The sectional area of new bone present at each time point ranged from 1.0 to 16.0 x 1O-4 mm2. This bone was woven and irregular in appearance. Tartrateresistant acid phosphatase and alkaline phosphatase activity was associated with cells in the marrow spaces and in the periosteum. Intense tartrateresistant acid phosphatase staining was located at the surface of the alveolar bone between CHA particles and bone (Fig 6A), and intense alkaline phosphatase staining was localized at the surface of the alveolar bone bordering the acid phosphatase activity (Fig 6B). In the body of the implant, fibrous tissue surrounded all CHA particles at all time points. Neither acute nor chronic inflammatory cells were present in the fibrous tissue surrounding the CHA. No bone or cartilage was seen in this fibrous tissue, and tartrate-resistant acid phosphatase and alkaline phosphatase stains were negative in this region. Composite onlays. Woven bone continuous with the outermost layer of alveolar bone was present at all time points. Within the body of the implant, each CHA particle was surrounded by fibrous tissue by day 7. Regions of chondrogenesis in day 14 specimens consisted of large chondroblasts surrounded by darkly staining metachromatic matrix. These regions, which were located between DBP particles not in close proximity to CHA, exhibited alkaline phosphatase staining. Induced bone was distributed throughout the implant at day 21
1072
TISSUE RESPONSE
FIGURE 5. Mandibular CHA onlay, day 7. Incisor can be seen at bottom center of photograph surrounded by alveolar bone. Of particular interest is the area of reactive bone (arrows) located between outermost layer of alveolar bone and implant. Spaces previously occupied by CHA contain debris and are surrounded by fibrous tissue and covered by gingival mucosa ttoluidine blue, original magnification X 40).
(Fig 7). The bone contained an abundance of osteoblasts with intense alkaline phosphatase activity. Occasional sites of osteoclast-mediated resorption were identified by tartrate-resistant acid phosphatase activity. Discrete periosteum was not distinguishable; induced bone appeared to be continuous with reactive bone from the mandible. Composite onlays generated more new bone formation at the mandibular site than did sham procedures or CHA onlays. The sectional area of new bone ranged from 2 to 32 x lop4 mm’. CALVARIAL SPECIMENS
Control. The calvariae were composed of a central marrow space enclosed by outer and inner tables of cortical bone. These tables exhibited a regular lamellar pattern parallel to the outer surface. Thin layers of fibrous tissue covered the scalp surface of the outer table and the brain surface of the inner table. No cartilage was present. Mild tartrate-
TO CHAiDBP IMPLANTS
resistant acid phosphatase and alkaline phosphatase activity was present in cells of the periosteum and endosteum. Sham. In most regions of sham-operated calvariae, the outer and inner tables of bone, their fibrous tissue covering, and the marrow spaces exhibited a histologic appearance similar to that of control calvariae. Reactive woven bone continuous with the superior surface of the outer table was occasionally present in day 10 specimens. There was no new bone present in the sham-operated calvariae of two animals; in the ones that showed a reaction to sham surgery, the region values were 1.8 x lop3 and 3.2 X 10m3 mm’. CHA onlays. In calvarial CHA onlays at day 7. the inner and outer tables and their surfaces and marrow spaces exhibited a histologic appearance similar to that of control specimens. In some regions, reactive bone formation extended from the underlying calvaria toward the implant (Fig 8). CHA onlay specimens exhibited more reactive bone formation than did sham-operated or control specimens. The sectional area of new bone ranged from 1.0 to 33.0 x lop4 mm*. Tartrate-resistant acid phosphatase and alkaline phosphatase activity was associated with cells in the endosteum. This enzyme activity was either reduced or absent by day 10. Occasional areas of direct contact between bone and CHA particles were observed. Within the body of the implant, CHA particles were surrounded by fibrous tissue by day 7. Neither inflammatory cells nor bone were seen in this region. Composite onlays. The histologic appearance of the outer and inner tables and marrow spaces of the calvarial composite onlay specimens was similar to that of control specimens. New bone continuous with the cortical bone of the outer table extended toward the implant. This woven bone contained alkaline phosphatase and tartrate-resistant acid phosphatase-positive cells in its marrow spaces. By day 10, enzyme activity were either reduced or absent. Within the body of the implant, fibrous tissue surrounded DBP and CHA particles. Inflammatory cells were not observed. There were islands of chondrogenesis in day 14 specimens. These areas consisted of large, alkaline phosphatase-positive chondroblasts surrounded by darkly staining metachromatic matrix, and were located between DBP particles not in close proximity to CHA. Induced bone was present at day 21 (Fig 9). The induced bone was often continuous with reactive bone located above the outer table. Composite onlays generated more total new bone formation at the calvarial site than did sham procedures or CHA onlays.
1073
PETTIS ET AL
FIGURE 6. A, Mandibular CHA onlay, day 7. Tartrate-resistant acid phosphatase (arrow) activity is located between the CHA particle and subjacent alveolar bone (original magnification X95). B, Mandibular CHA onlay, day 7. Alkaline phosphatase activity (arrows) is located near the surface of subjacent alveolar bone (AB) adjacent to tartrate-resistant acid phosphatase activity (alkaline phosphatase, original magnification X95).
The amount of new bone ranged from 3.0 to 290 x 10P4 mm’. Discussion Ceramic hydroxyapatite elicited a benign response in all sites. The absence of acute inflammation may be due to the smooth, rounded particle shape as well as the bioinert chemical nature of the CHA. The results of this study indicate that CHA is not osteoinductive. Bone was not induced extraskeletally; no cartilage or bone formation was observed in any of the subcutaneously implanted CHA specimens. Alkaline phosphatase activity, which serves as a marker for bone formation, was absent. Although CHA did not induce osteogenesis, it did promote osteoconduction when used as an onlay, that is, reactive bone was produced at the surface of the adjacent bone. Enzyme activity localization provides information regarding the possible mechanism of initiation of osteoconduction. Figure 6A illustrates the tar-
trate-resistant acid phosphatase activity that was frequently observed between the particles and underlying alveolar bone. This activity represents the activation of resorption initiated beneath the particles. Alkaline phosphatase activity (Fig 6B) was adjacent to regions with tar&ate-resistant acid phosphatase. The intimate coupling between bone resorption and formation I3314 has been previously reported. Whether this relationship is regulated by chemical mediators, piezoelectric signals, surface charge phenomena, or other factors is not revealed by these studies. Sham-operated controls were included to evaluate the role of surgical trauma in the initiation of bone resorption. Trauma from the excision of calvarial periosteum elicited reactive bone formation in some cases. This type of bone formation may account for a small portion of the CHA-elicited osteoconduction in calvarial onlays. Surgical preparation of mandibular sites in the sham group, however. did not appear to stimulate the osteoconduc-
FIGURE 7. Mandibular composite onlay, day 21. Arrows denote areas of induced bone contiguous with particles of CHA and united to the alveolar bone below (toluidine blue. original magnification X95). FIGURE 8. Calvarial CHA onlay, day 14. Arrows denote reactive bone formation stimulated by CHA onlays. Note areas of bone to CHA contact (toluidine blue, original magnification X95).
1074
TISSUE RESPONSE TO CHA/DBP IMPLANTS
The ideal synthetic implant material for reconstruction of the atrophic mandibular ridge should provide a firm consistency in the surgical area a relatively short time postoperatively. ” The composite implant may provide a means of achieving this ideal by combining the phenomenon of induced osteogenesis with the structural support provided by ceramics. Acknowledgments
FIGURE 9. Calvarial composite onlay, day 21. New bone (arrows) has engulfed DBP particles and partially surrounded CHA particles in the center of the slide (toluidine blue, original magnification X95)
tive bone formation observed in calvarial sham specimens. The decrease in both tat-u-ate-resistant acid phosphatase and alkaline phosphatase activity in the marrow spaces of CHA-elicited bone at day 10 suggests that reactive bone formation occurred in a brief time frame. The absence of an increase in the amount of new bone present after day 7 supports this hypothesis. Therefore, the differences in the amount of new bone present over time may reflect sampling variations. Long-term studies are needed to determine whether additional bone formation, remodeling, or resorption occur at times later than those studied herein. Subcutaneous composite implant specimens showed chondrogenesis by day 14 and did not exhibit bone formation until day 21. More frequent sampling is needed to establish whether CHA significantly delays DBP-induced osteoinduction. The mixture used for our study was obtained by incorporating a small amount of a collagen suspension into the 4:l mixture. The collagen appeared to be effective in carrying the components. Results from this study support those of Kent et al, who reported that CHA onlays stimulated only minimum ingrowth of reactive bone into the augmented mandibular ridges of dogs over 52 weeks.16 Therefore, both studies indicate that maximum contour augmentation could not be achieved solely by CHA-elicited osteoconduction. Incorporation of fresh autogenous bone into CHA before implantation increased the amount of bone that formed in the augmented ridge; however, significant ridge height was lost by 12 weeks.16 When a combination of demineralized bone matrix and CHA was used for mandibular ridge augmentation, ridge height was maintained; however, the ridges exhibited bone formation only after 26 weeks.
This study was done in partial fulfillment of the requirements for the Master of Medical Sciences Degree in Oral Biology at Harvard University. The authors wish to express their appreciation for the expert technical assistance provided by Sandy Wilcon.
References 1. Baker RD, Terry BC, Davis WH, et al: Long-term results of alveolar ridge augmentation. J Oral Surg 37:486, 1979 2. Kelly JF, Friedlander GE: Preprosthetic bone graft augmentation with allogeneic bone: A preliminary report. J Oral Surg 35~268, 1977 3. Kent JN, Quinn JH, Zide MF, et al: Correction of alveolar ridge deficiencies with nonresorbable hydroxyapatite. J Am Dent Assoc 105993, 1982 4. Jarcho M: Calcium phosphate ceramics as hard tissue prosthetics. Clin Orthop 157:259, 1981 5. Urist MR: Bone: Formation by autoinduction. Science 150:893, 1965 6. Glowacki J, Murray JE, Kaban LB, et al: Application of the biological principle of induced osteogenesis for craniofacial defects. Lancet i:959, 1981 7. Sonis SS, Kaban LB, Glowacki J: Clinical trial of demineralized bone powder in the treatment of periodontal defects. J Oral Med 38: 117, 1983 8. Glowacki J, Altobelli D, Mulliken JB: Fate of mineralized and demineralized osseous implants in cranial defects. Calcif Tissue Int 33:71, 1981 9. Kaban LB, Glowacki J: Augmentation of rat mandibular ridge with demineralized bone implants. J Dent Res 63:998, 1984 10. Reddi AH, Huggins C: Biochemical sequences in the transformation of normal tibroblasts in adolescent rats. Proc Nat1 Acad Sci USA 69: 1601, 1972 11. Glowacki J, Mulliken JB: Demineralized bone implants. Clin Plast Surg 12:233, 1985 12. Fallon MD. Teitelbaum SL: A simple procedure for the rapid histologic diagnosis of metabolic bone disease. Calcif Tissue Int 33:281, 1981 13. Baron R, Vignery A, Horowitz M: Lymphocytes, macrophages and the regulation of bone remodeling, in Bone and Mineral Research, Annual 2. New York, Elsevier, 1983, pp 175-243 14. Kahn AJ, Fallon MD, Teitelbaum SL: Structure-function relationships in bone: an examination of events at the cellular level, in Bone and Min Research, Annual 2. New York, Elsevier, 1983, pp 125-173 15. Block MS, Kent JN, Ardoin RC, et al: Mandibular augmentation in dogs with hydroxyapatite combined with demineralized bone. J Oral Maxillofac Surg 45:414, 1987 16. Block MS, Kent JN: Healing of mandibular ridge augmentations using hydroxyapatite with and without autogenous bone in dogs. J Oral Maxillofac Surg 43:3, 1985 17. Griffith GR: New hydroxyapatite ceramic materials: Potential use for bone induction and alveolar ridge augmentation. J Prosthet Dent 53:109, 1985
Surg
48:10@3-1074,1990
Tissue Response to Composite Ceramic Hydroxyapa titel Demineralized Bone Implants GAIL Y. PETTIS, DDS, MMEDSC,* AND
JULIE
LEONARD
GLOWACKI,
B. KABAN,
DMD,
MD,t
PHD$
This study evaluated the tissue reactions to two materials: ceramic hydroxyapatite (CHA), and a composite material of demineralized bone powder (DBP) and CHA (ratio of 4:l) in a collagen vehicle. The materials were tested in a subcutaneous pocket, a mandibular onlay, and in a calvarial onlay model. Specimens were evaluated histologically at 7, 10, 14, and 21 days postimplantation. Ceramic hydroxyapatite, implanted subcutaneously, elicited a fibrous response with minimal inflammation, but did not induce bone formation. In specimens of subcutaneously implanted composite material, induced bone was evident in association with the DBP. In CHA onlay specimens, there was a small amount of reactive bone extending from the host bone into the implant. In composite onlays, bone filled the entire body of the implant. The results of this study indicate that CHA particles were not osteoinductive in heterotopic sites and that osteoconductive ingrowth was minimal in onlays. Bone was induced by DBP even when mixed with CHA particles and implanted in subcutaneous and intraosseous sites. It was concluded that composite implants may provide a means of combining the osteoinductive properties of DBP with the bulk and structural support of osteoconductive CHA particles.
The atrophic mandibular ridge provides one of the most difficult challenges in the field of reconstructive surgery. Conventional treatment involves augmentation with onlays of autogenous bone. These grafts, however, undergo progressive, uncontrolled resorption during healing.’ Donor site morbidity is an additional drawback to the use of fresh autogenous bone.2 These disadvantages have stimulated the development of alternative implant
materials, such as particulate ceramic calcium phosphates.3 In addition to eliminating donor site morbidity, these materials provide immediate firmness and are available in nonresorbable forms.4 A disadvantage to the use of particulate ceramics is their tendency to migrate through the surrounding connective tissue.3 Furthermore, the nature of the tissue response to these materials is inadequately understood. Particulate ceramics provide a suitable substrate for bone deposition, a property known as osteoconduction. Controversy exists, however, regarding whether osteoconduction reaches the height of the ridge augmented with ceramics. Demineralized bone has been demonstrated as an alternative material for several types of osseous reconstruction.5-7 The mechanism of healing with demineralized bone matrix is osteoinduction. The matrix stimulates phenotypic conversion of connective tissue cells into chondroblasts with endochondral bone formation.839 Therefore, extraskeletal implantation sites may be used to distinguish between inductive and conductive mechanisms of healing.”
* Assistant Professor, Department of Orthodontics, College of Dentistry, University of Tennessee, Memphis, TN. t Professor and Chairman, Department of Oral and Maxillofacial Surgery, University of California School of Dentistry, San Francisco, CA. $ Associate Professor, Department of Orthopedic Surgery, Harvard Medical School, Brigham and Women’s Hospital, Boston, MA. Address correspondence and reprint requests to Dr Pettis: Department of Orthodontics, University of Tennessee, College of Dentistry, 875 Union Ave, Memphis, TN 38163. 01990 geons
American
0278-2391/90/481
Association
of Oral and Maxillofacial
Sur-
O-001 0$3.00/O
1068
1069
PETTIS ET AL
Kaban and Glowacki’ developed a rat mandibular model to evaluate demineralized bone matrix and other bone-derived graft materials for use in mandibular ridge augmentation. Various materials were placed supraperiosteally in the edentulous area located between the incisor and first molar. Unlike mineral-containing bone particles, demineralized bone matrix induced a mass of bone that was firmly united to the ridge by the 14th day after implantation. However, the practical drawback to the clinical use of demineralized bone implants is that, unlike ceramics, they do not provide immediate rigidity. In this study, we evaluated tissue reactions to particulate CHA and to a composite of demineralized bone powder and CHA in three sites. Tissue responses to implants of DBP have been well documented in animal models8.9*“; these sites were also implanted with CHA alone to determine the portion of the healing response attributable to the ceramic. Subcutaneous pouches were used for positive identification of an osteoinductive process. Osteoconduction was observed in mandibular onlays. The calvarial onlay site provided the additional benefit of direct access for complete removal of the periosteum. Methods and Materials
PREPARATIONOF IMPLANT MATERIALS Particulate CHA (Calcitite 2040) was obtained from Calcitek, Inc (San Diego, CA). The particles were dense spheres with diameters ranging from 425 to 850 km. Approximately 0.2 cm2 was used for each implant. Demineralized bone powder was prepared for composite implants according to the protocol of Glowacki and Mulliken.” Briefly, diaphyses of long bones were obtained from Charles River rats. After the bones were scraped free of soft tissue and marrow and thoroughly washed in cold, deionized water, they were extracted with ethanol followed by anhydrous ethyl ether. The resulting dehydrated material was pulverized in a liquid nitrogen impacting mill (Spex, Metuchen, NJ). Pulverized bone particles were sieved into fractions according to particle size. Particles in the range of 75 to 250 pm were demineralized with 0.5 mol/L hydrochloric acid (50 mL per g bone) for 3 hours at room temperature. Acid and liberated minerals were washed away with cold, deionized water (500 mL per g) until the pH of the wash matched the pH of the water. After extensive washings with water, the DBP was extracted with absolute ethanol followed by anhydrous ether. The DBP was mixed with CHA in a 4: 1 ratio by volume. The mixture (0.2 mL) was moistened with
approximately 0.1 mL of Zyderm collagen (Collagen Corp, Palo Alto, CA). SURGICAL PROCEDURES Both CHA and the composite mixtures were implanted in subcutaneous pockets and used as onlays for the mandible and calvarium. Sham operations were performed in onlay sites so that histologic changes due to surgical trauma could be evaluated. In addition, mandibles and calvariae were obtained from intact, age-matched animals for use as controls. All specimens were obtained at 7, 10, 14, and 21 days postoperatively. Site I: Subcutaneous
Pouches
Twelve 28-day-old male Charles River rats were anesthetized by inhalation of methoxyflurane. After the surgical area was shaved and swabbed with alcohol, bilateral ventral incisions were made on either side of the midline at the level of the diaphragm. Narrow cephalad pockets were created by blunt dissection and approximately 0.2 cm2 of implant material was deposited at the end of the pocket, 1 cm away from the incision. The incision was closed with a 9-mm wound clip. Site 2: Supraperiosteal Mandibular Onlays Twenty-four 28-day-old male Charles River rats were sedated and anesthetized by inhalation of methoxyflurane and intraperitoneal injections of Brevital (5 mg/kg; Eli Lilly, Indianapolis, IN). Lidocaine (2% with epinephrine l:lOO,OOO)was infiltrated into the submental region bilaterally. After the surgical area was shaved and swabbed with alcohol, a horizontal incision was made in the submental triangle just posterior to the mandibular symphysis. Both horizontal rami were exposed and a submucosal pocket created anteroposteriorly on the superior aspect of the edentulous region between incisor and molars, leaving the periosteum attached to bone. Test materials were implanted and deep closure achieved with 4-O chromic catgut suture. The skin incision was closed with two 9-mm wound clips. These animals received a diet of milled standard rat chow for 2 weeks following surgery. Specimens were harvested by excision through the body of the mandible in the molar region. Sham operations were performed on four rats. Anesthesia and soft-tissue dissection were performed as above; however, no implants were placed. These animals received a diet of milled standard rat chow for 2 weeks.
1070
TISSUE RESPONSETO CHA/DBPIMPLANTS
Site 3: Calvarial Onlays
Twenty-four 28-day-old male Charles River rats were sedated and anesthetized by inhalation of methoxyflurane and intraperitoneal injections of Brevital(5 mg/kg). After shaving and swabbing with alcohol, lidocaine (2% with epinephrine 1: 100,000) was infiltrated into the surgical area. A coronal incision was made at the level of the ears. The scalp was reflected anteriorly and periosteum removed from the bone. Implant material was placed over the sagittal suture between the frontal and occipital sutures. Incisions were closed with interrupted 4-O nylon suture. Sham operations included anesthesia and soft-tissue dissection, and complete removal of the periosteum. These procedures were performed on four rats. HISTOLOGIC METHODS After they were killed, all specimens were immediately fixed in cold 2% paraformaldehyde in 0.1 mol/L cacodylate buffer (pH 7.4) for 24 hours. Mandibular and calvarial specimens were demineralized at 4°C with frequent changes of the demineralizing fluid, which was either 7.5% EDTA mixed in 0.1 mol/L cacodylate or a solution consisting of 0.05 mol/L of formic acid and 0.03 mol/L sodium citrate. After demineralization for 3 weeks, specimens were rinsed thoroughly with cold, deionized water and embedded in glycol methacrylate embedding media. Three-micrometer histologic sections were made as follows, with a Jung Model 1140 Autocut microtome (Reichert-Jung, Vienna, Austria) fitted with a Carborundum knife. 1. Subcutaneous implants: The section was taken at approximately one-third to one-half the depth of the implant. 2. Mandibular implants: The section was taken in the coronal plane, at the anterior-posterior midpoint of the implant. 3. Calvarial implants: Each calvarial specimen was divided into two portions along the sagittal plane. Coronal sections were made approximately 1 mm from the cut edge of each half. Sections were stained with toluidine blue. Additional sections were stained for tarn-ate-resistant acid phosphatase activity (a marker of osteoclast activity) and for alkaline phosphatase activity, which served as a marker for osteoblast activity.12 The following features were quantitated in the sections: 1) the total area of the implant, 2) the area occupied by ceramic particles, and 3) the area of new bone within or adjacent to the implant. A mag-
nification conversion units of mm*.
factor was applied to obtain Results
SUBCUTANEOUS IMPLANTS
CHA. By day 7, each particle was enclosed by layers of fibrous connective tissue (Fig 1). With the exception of an occasional multinucleated cell, neither acute nor chronic inflammatory cells were evident at any time point. This indicated that acute inflammation that may have been associated with the surgical procedure or with CHA was resolved by day 7 and that CHA did not elicit a foreign-body reaction. No cartilage or bone was observed in any of the specimens. Staining for tartrate-resistant acid phosphatase and alkaline phosphatase activities was negative. Composite. Figure 2 demonstrates the distribution of CHA and DBP particles in the preimplanted composite material. Subcutaneous composite implant specimens were highly cellular with fibrous connective tissue surrounding each particle at day 7. By day 14, islands of induced cartilage were present between the DBP particles, but not in close proximity to CHA (Fig 3). By day 21, rows of osteoblasts and occasional foci of osteoclast-mediated bone resorption could be identified (Fig 4). Alkaline phosphatase activity was associated with the osteoblasts, and positive staining for tartrate-resistant acid phosphatase was identified in the osteoclasts, especially at their surfaces adjacent to bone. MANDIBULAR SPECIMENS
Control. Coronal sections at the anterior-posterior midpoint of the mandibular implantation site
FIGURE 1. Subcutaneous CHA implant, day 7. “CHA” denotes region previously occupied by ceramic particles. (Particles of ceramic were lost in the sectioning procedure. Ceramic remnants remain in these spaces.) (Toluidine blue, original magnification x 50.)
PETTIS ET AL
FIGURE 2. Preimplanted composite material. “CHA” denotes spaces previously occupied by the ceramic implant. DBPs are distributed throughout. with CHA interspersed ttoluidine blue, original magnification x38).
demonstrated the normal anatomy of the rat mandible. The incisor was located in the center of histologic cross sections through the control mandibles. The outermost layer of the alveolus consisted of lamellar bone with a regular circumferential pattern of organization. This bone was surrounded by thin, fibrous periosteum. Tartrate-resistant acid phosphatase and alkaline phosphatase activity indicated remodeling at the surface of this bone. Immediately superior to the periosteum was a thin layer of muscle covered by gingival mucosa and epithehum. Sham. The histologic appearance of the incisor and its supporting structures in the sham-operated specimens was identical to that of the control specimens with the exception of an occasional area of tissue separation between periosteum and muscle or gingival mucosa. These results indicate that any acute inflammation that may have occurred due to
FIGURE 3. Subcutaneously implanted composite material, day 14. Spaces previously occupied by CHA are located at the periphery of the field, with DBP particles in the center. Arrows denote areas of chondrogenesis located between DBP particles (toluidine blue, original magnification x38).
1071
FIGURE 4. Subcutaneously implanted composite material, day 21. CHA particle spaces can be seen at the periphery of the field. Arrows denote area of induced bone. (toluidine blue, original magnification x38).
surgical manipulation had resolved by day 7. No reactive woven bone was present. CHA onlays. The histologic appearance of the incisor and the bone surrounding it was similar to that of control specimens. However, at all time points, there were areas of newly formed bone continuous with the outermost layer of alveolar bone defined by the control specimens (Fig 5). The sectional area of new bone present at each time point ranged from 1.0 to 16.0 x 1O-4 mm2. This bone was woven and irregular in appearance. Tartrateresistant acid phosphatase and alkaline phosphatase activity was associated with cells in the marrow spaces and in the periosteum. Intense tartrateresistant acid phosphatase staining was located at the surface of the alveolar bone between CHA particles and bone (Fig 6A), and intense alkaline phosphatase staining was localized at the surface of the alveolar bone bordering the acid phosphatase activity (Fig 6B). In the body of the implant, fibrous tissue surrounded all CHA particles at all time points. Neither acute nor chronic inflammatory cells were present in the fibrous tissue surrounding the CHA. No bone or cartilage was seen in this fibrous tissue, and tartrate-resistant acid phosphatase and alkaline phosphatase stains were negative in this region. Composite onlays. Woven bone continuous with the outermost layer of alveolar bone was present at all time points. Within the body of the implant, each CHA particle was surrounded by fibrous tissue by day 7. Regions of chondrogenesis in day 14 specimens consisted of large chondroblasts surrounded by darkly staining metachromatic matrix. These regions, which were located between DBP particles not in close proximity to CHA, exhibited alkaline phosphatase staining. Induced bone was distributed throughout the implant at day 21
1072
TISSUE RESPONSE
FIGURE 5. Mandibular CHA onlay, day 7. Incisor can be seen at bottom center of photograph surrounded by alveolar bone. Of particular interest is the area of reactive bone (arrows) located between outermost layer of alveolar bone and implant. Spaces previously occupied by CHA contain debris and are surrounded by fibrous tissue and covered by gingival mucosa ttoluidine blue, original magnification X 40).
(Fig 7). The bone contained an abundance of osteoblasts with intense alkaline phosphatase activity. Occasional sites of osteoclast-mediated resorption were identified by tartrate-resistant acid phosphatase activity. Discrete periosteum was not distinguishable; induced bone appeared to be continuous with reactive bone from the mandible. Composite onlays generated more new bone formation at the mandibular site than did sham procedures or CHA onlays. The sectional area of new bone ranged from 2 to 32 x lop4 mm’. CALVARIAL SPECIMENS
Control. The calvariae were composed of a central marrow space enclosed by outer and inner tables of cortical bone. These tables exhibited a regular lamellar pattern parallel to the outer surface. Thin layers of fibrous tissue covered the scalp surface of the outer table and the brain surface of the inner table. No cartilage was present. Mild tartrate-
TO CHAiDBP IMPLANTS
resistant acid phosphatase and alkaline phosphatase activity was present in cells of the periosteum and endosteum. Sham. In most regions of sham-operated calvariae, the outer and inner tables of bone, their fibrous tissue covering, and the marrow spaces exhibited a histologic appearance similar to that of control calvariae. Reactive woven bone continuous with the superior surface of the outer table was occasionally present in day 10 specimens. There was no new bone present in the sham-operated calvariae of two animals; in the ones that showed a reaction to sham surgery, the region values were 1.8 x lop3 and 3.2 X 10m3 mm’. CHA onlays. In calvarial CHA onlays at day 7. the inner and outer tables and their surfaces and marrow spaces exhibited a histologic appearance similar to that of control specimens. In some regions, reactive bone formation extended from the underlying calvaria toward the implant (Fig 8). CHA onlay specimens exhibited more reactive bone formation than did sham-operated or control specimens. The sectional area of new bone ranged from 1.0 to 33.0 x lop4 mm*. Tartrate-resistant acid phosphatase and alkaline phosphatase activity was associated with cells in the endosteum. This enzyme activity was either reduced or absent by day 10. Occasional areas of direct contact between bone and CHA particles were observed. Within the body of the implant, CHA particles were surrounded by fibrous tissue by day 7. Neither inflammatory cells nor bone were seen in this region. Composite onlays. The histologic appearance of the outer and inner tables and marrow spaces of the calvarial composite onlay specimens was similar to that of control specimens. New bone continuous with the cortical bone of the outer table extended toward the implant. This woven bone contained alkaline phosphatase and tartrate-resistant acid phosphatase-positive cells in its marrow spaces. By day 10, enzyme activity were either reduced or absent. Within the body of the implant, fibrous tissue surrounded DBP and CHA particles. Inflammatory cells were not observed. There were islands of chondrogenesis in day 14 specimens. These areas consisted of large, alkaline phosphatase-positive chondroblasts surrounded by darkly staining metachromatic matrix, and were located between DBP particles not in close proximity to CHA. Induced bone was present at day 21 (Fig 9). The induced bone was often continuous with reactive bone located above the outer table. Composite onlays generated more total new bone formation at the calvarial site than did sham procedures or CHA onlays.
1073
PETTIS ET AL
FIGURE 6. A, Mandibular CHA onlay, day 7. Tartrate-resistant acid phosphatase (arrow) activity is located between the CHA particle and subjacent alveolar bone (original magnification X95). B, Mandibular CHA onlay, day 7. Alkaline phosphatase activity (arrows) is located near the surface of subjacent alveolar bone (AB) adjacent to tartrate-resistant acid phosphatase activity (alkaline phosphatase, original magnification X95).
The amount of new bone ranged from 3.0 to 290 x 10P4 mm’. Discussion Ceramic hydroxyapatite elicited a benign response in all sites. The absence of acute inflammation may be due to the smooth, rounded particle shape as well as the bioinert chemical nature of the CHA. The results of this study indicate that CHA is not osteoinductive. Bone was not induced extraskeletally; no cartilage or bone formation was observed in any of the subcutaneously implanted CHA specimens. Alkaline phosphatase activity, which serves as a marker for bone formation, was absent. Although CHA did not induce osteogenesis, it did promote osteoconduction when used as an onlay, that is, reactive bone was produced at the surface of the adjacent bone. Enzyme activity localization provides information regarding the possible mechanism of initiation of osteoconduction. Figure 6A illustrates the tar-
trate-resistant acid phosphatase activity that was frequently observed between the particles and underlying alveolar bone. This activity represents the activation of resorption initiated beneath the particles. Alkaline phosphatase activity (Fig 6B) was adjacent to regions with tar&ate-resistant acid phosphatase. The intimate coupling between bone resorption and formation I3314 has been previously reported. Whether this relationship is regulated by chemical mediators, piezoelectric signals, surface charge phenomena, or other factors is not revealed by these studies. Sham-operated controls were included to evaluate the role of surgical trauma in the initiation of bone resorption. Trauma from the excision of calvarial periosteum elicited reactive bone formation in some cases. This type of bone formation may account for a small portion of the CHA-elicited osteoconduction in calvarial onlays. Surgical preparation of mandibular sites in the sham group, however. did not appear to stimulate the osteoconduc-
FIGURE 7. Mandibular composite onlay, day 21. Arrows denote areas of induced bone contiguous with particles of CHA and united to the alveolar bone below (toluidine blue. original magnification X95). FIGURE 8. Calvarial CHA onlay, day 14. Arrows denote reactive bone formation stimulated by CHA onlays. Note areas of bone to CHA contact (toluidine blue, original magnification X95).
1074
TISSUE RESPONSE TO CHA/DBP IMPLANTS
The ideal synthetic implant material for reconstruction of the atrophic mandibular ridge should provide a firm consistency in the surgical area a relatively short time postoperatively. ” The composite implant may provide a means of achieving this ideal by combining the phenomenon of induced osteogenesis with the structural support provided by ceramics. Acknowledgments
FIGURE 9. Calvarial composite onlay, day 21. New bone (arrows) has engulfed DBP particles and partially surrounded CHA particles in the center of the slide (toluidine blue, original magnification X95)
tive bone formation observed in calvarial sham specimens. The decrease in both tat-u-ate-resistant acid phosphatase and alkaline phosphatase activity in the marrow spaces of CHA-elicited bone at day 10 suggests that reactive bone formation occurred in a brief time frame. The absence of an increase in the amount of new bone present after day 7 supports this hypothesis. Therefore, the differences in the amount of new bone present over time may reflect sampling variations. Long-term studies are needed to determine whether additional bone formation, remodeling, or resorption occur at times later than those studied herein. Subcutaneous composite implant specimens showed chondrogenesis by day 14 and did not exhibit bone formation until day 21. More frequent sampling is needed to establish whether CHA significantly delays DBP-induced osteoinduction. The mixture used for our study was obtained by incorporating a small amount of a collagen suspension into the 4:l mixture. The collagen appeared to be effective in carrying the components. Results from this study support those of Kent et al, who reported that CHA onlays stimulated only minimum ingrowth of reactive bone into the augmented mandibular ridges of dogs over 52 weeks.16 Therefore, both studies indicate that maximum contour augmentation could not be achieved solely by CHA-elicited osteoconduction. Incorporation of fresh autogenous bone into CHA before implantation increased the amount of bone that formed in the augmented ridge; however, significant ridge height was lost by 12 weeks.16 When a combination of demineralized bone matrix and CHA was used for mandibular ridge augmentation, ridge height was maintained; however, the ridges exhibited bone formation only after 26 weeks.
This study was done in partial fulfillment of the requirements for the Master of Medical Sciences Degree in Oral Biology at Harvard University. The authors wish to express their appreciation for the expert technical assistance provided by Sandy Wilcon.
References 1. Baker RD, Terry BC, Davis WH, et al: Long-term results of alveolar ridge augmentation. J Oral Surg 37:486, 1979 2. Kelly JF, Friedlander GE: Preprosthetic bone graft augmentation with allogeneic bone: A preliminary report. J Oral Surg 35~268, 1977 3. Kent JN, Quinn JH, Zide MF, et al: Correction of alveolar ridge deficiencies with nonresorbable hydroxyapatite. J Am Dent Assoc 105993, 1982 4. Jarcho M: Calcium phosphate ceramics as hard tissue prosthetics. Clin Orthop 157:259, 1981 5. Urist MR: Bone: Formation by autoinduction. Science 150:893, 1965 6. Glowacki J, Murray JE, Kaban LB, et al: Application of the biological principle of induced osteogenesis for craniofacial defects. Lancet i:959, 1981 7. Sonis SS, Kaban LB, Glowacki J: Clinical trial of demineralized bone powder in the treatment of periodontal defects. J Oral Med 38: 117, 1983 8. Glowacki J, Altobelli D, Mulliken JB: Fate of mineralized and demineralized osseous implants in cranial defects. Calcif Tissue Int 33:71, 1981 9. Kaban LB, Glowacki J: Augmentation of rat mandibular ridge with demineralized bone implants. J Dent Res 63:998, 1984 10. Reddi AH, Huggins C: Biochemical sequences in the transformation of normal tibroblasts in adolescent rats. Proc Nat1 Acad Sci USA 69: 1601, 1972 11. Glowacki J, Mulliken JB: Demineralized bone implants. Clin Plast Surg 12:233, 1985 12. Fallon MD. Teitelbaum SL: A simple procedure for the rapid histologic diagnosis of metabolic bone disease. Calcif Tissue Int 33:281, 1981 13. Baron R, Vignery A, Horowitz M: Lymphocytes, macrophages and the regulation of bone remodeling, in Bone and Mineral Research, Annual 2. New York, Elsevier, 1983, pp 175-243 14. Kahn AJ, Fallon MD, Teitelbaum SL: Structure-function relationships in bone: an examination of events at the cellular level, in Bone and Min Research, Annual 2. New York, Elsevier, 1983, pp 125-173 15. Block MS, Kent JN, Ardoin RC, et al: Mandibular augmentation in dogs with hydroxyapatite combined with demineralized bone. J Oral Maxillofac Surg 45:414, 1987 16. Block MS, Kent JN: Healing of mandibular ridge augmentations using hydroxyapatite with and without autogenous bone in dogs. J Oral Maxillofac Surg 43:3, 1985 17. Griffith GR: New hydroxyapatite ceramic materials: Potential use for bone induction and alveolar ridge augmentation. J Prosthet Dent 53:109, 1985
