Hydroxyapatite I. Basic
Cement
Chemistry and Histologic Properties
Peter D. Costantino, MD; Craig D. Friedman, MD; Kent Jones, MD; Lawrence C. Chow, PhD; Harold J. Pelzer, DDS, MD; George A. Sisson, Sr, MD
Hydroxyapatite cement is a unique calphosphate preparation that can be shaped intraoperatively and sets in vivo to an implant composed of microporous hydroxyapatite. The histologic response to this cement was evaluated by implanting \s=b\
cium
disks made of this material within the heads of nine cats. Three sets of 12 hydroxyapatite cement disks were produced containing 0%, 10%, and 20% macropores by volume, respectively. The disks were implanted subcutaneously, intramuscularly, above the periosteum of the skull, and directly onto the surface of the calvarium. Each macropore percentage was represented in each tissue plane, and animals were killed up to 9 months postoperatively. There were no toxic reactions, implants extruded, or wound infections. Histologic examination of the implant\p=m-\soft-tissue interfaces revealed a
Accepted for publication November 2,1990. From the Department of Otolaryngology\p=m-\Head and Neck Surgery, Northwestern University Medical School, Chicago, Ill (Drs Costantino, Jones, Pelzer, and Sisson); Section of Otolaryngology\p=m-\ Head and Neck Surgery, Yale University Medical School, New Haven, Conn (Dr Friedman); and the
American Dental Association Health Foundation, Paffenbarger Research Institute at the National Institute of Standards and Technology, Gaithersburg, Md (Dr Chow). Presented at the Fifth International Symposium on Facial Plastic Reconstructive Surgery of the Head and Neck, June 27,1989, Toronto, Ontario. Reprint requests to Section of Otolaryngology\p=m-\ Head and Neck Surgery, Yale University School of 333 Cedar St, New Haven, Medicine, PO Box CT 06510 (Dr Friedman).
3333,
inflammatory response without foreign body reaction. The disks were resorbed over time in direct proportion to their macropore content (surface areas) in all groups except for those disks placed directly onto the surface of the calvarium below the periosteum. In this group, numerous foci of bone formed at the skull-implant interface, with variable replacement of the deep surface of these implants by bone. Implant replacement by bone is postulated to occur through a combination of implant resorption coupled with osteoconduction. Based on these properties, hydroxyapatite cement may prove useful when applied to the reconstruction of non\p=m-\stress-bearing transient
skeletal tissue.
(Arch Otolaryngol Head 117:379-384)
Neck
Surg. 1991;
Alloplastic implants
have been used craniofacial skeleton since 1600, when Fallopius to reconstruct the
implanted
a
gold plate
to
repair
a
cranial defect.1 Since that time, con¬ tour reconstruction of the non-stressbearing, craniofacial skeleton has con¬ tinued to be a technical and materials problem. Metals are difficult to shape and are hampered by problems such as infection and corrosion.23 Polymers such as silicone, Proplast, or methyl-
methacrylate are encapsulated by scar resulting in significant rates of implant infection and/or extrusion.4"* Biologic materials, such as autogenous bone grafts, may cause donor site mor¬ bidity, suffer from significant postim¬ plantation résorption, and are difficult to accurately conform to skeletal de¬
tissue
fects.9"11 At present, there is no ideal natural or synthetic material for con¬ tour reconstruction of the craniofacial
skeleton. Of those alloplastic materials used to augment and replace the facial skele¬ ton, the most promising and well tol¬ erated are calcium phosphate-based compounds.'12"14 The majority of the human skeleton is composed of calcium phosphate in the form of hydroxyapa¬ tite (HA), and processing techniques developed during the early 1970s have provided HA preparations with suffi¬ cient structural strength to serve as bone substitutes.1" Since the mid1970s, various preformed HA prepara¬ tions have been used in a variety of clinical applications within medicine and dentistry.1'1"18 The applicability of these HA preparations, however, has been limited, because these implants had to be preformed as a hard materi¬
al.18
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
Fig 1. Solubility phase diagram for various cal¬ cium phosphates at 37°C. OHAp indicates hydroxyapatite; ß-TCP, beta-tricalcium phos¬ phate; TTCP, tetracalcium phosphate; OCP, octacalcium phosphate; DDCP, dihydrate dicalcium phosphate; and ADCP, anhydrous dicalcium phosphate. —
Recently, a new type of calcium phosphate cement that sets to HA in
vivo has been developed19 at the Paffenbarger Research Center of the American Dental Association Health Foundation (ADAHF) located at the National Institutes of Standards and Technology, Gaithersburg, Md. This material is referred to as HA cement (HAC) and is particularly unique in that it can be applied intraoperatively
paste and, subsequently, sets to a structurally stabile implant composed of microporous HA. All other forms of HA, to our knowledge, must be pre¬ formed as a hard material prior to implantation. The purpose of this study was to test the histologie response to HAC in a large animal model. Application of as a
HAC to frontal sinus obliteration and reconstruction has also been per¬ formed and is described in part II20 of this study. The hypotheses tested herein were that HAC would be highly tissue compatible, would not be resorbable, and would demonstrate properties potentially suitable for use in the augmentation of nonstress-bearing areas of the craniofacial skeleton. BACKGROUND Biologic Characteristics of HA Compounds
The HA is
general chemical composition of [Cairj(P04)6(OH)2], and it essen¬
tially represents interlinked calcium phosphate molecules. Hydroxyapatite
is the primary mineral component of teeth and bone and exists in bone in equilibrium with other calcium phos¬ phate salts. At a pH of 7.4, it is the least soluble of the naturally occurring calcium phosphate salts (Fig 1). Its stable macromolecular structure, along with its low solubility at physiologic pH,1'19 results in a relative resistance to résorption in vivo, depending on the method of preparation. All forms of HA are noteworthy for their excellent biocompatibility. Hydroxyapatite does not cause a foreign body giant cell reaction, a sustained inflammatory response, toxic reac¬ tions, or an increase in serum calcium or phosphate levels. The most impor¬ tant characteristic of HA compounds is their interaction with osseous tissue. Some studies have claimed that the presence of HA compounds within the body induces osteogenesis (de novo formation of new bone), but definitive studies21 using bone marrow and HA in implanted Millipore filters showed no osteogenic effect by HA in the absence of bone marrow cells.21 Although HA is not osteogenic, it is osteoconductive, in that it can serve as a scaffold on which bone can grow. If macropores (200 to 300 µ ) are fabri¬ cated into an HA implant, bone grows into the macropores. "',22"24 When im¬ planted into the cortex of long bones in dogs, preformed porous ceramic HA blocks becomes ingrown with osteoid tissue.1" Implanted blocks of HA can show a faster rate of bone ingrowth than can bone grafts,23 and similar studies have also shown that cartilage is able to grow into porous HA im¬ plants. Despite a lack of osteogenesis, a direct chemical bond forms between the implant and bone without interven¬ ing fibrous tissue. This bonding of im¬ plant to bone is referred to as
osseointegration.18'22
Types of HA:
Ceramic and Nonceramic
Hydroxyapatite can be fabricated in ceramic and nonceramic forms. Since the mid-1970s, HA has been used clini¬ cally only as a ceramic preparation. The ceramic forms of HA are heated to fuse individual HA crystals to each other through a process called sinter¬ ing. After HA crystals are fabricated
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
a pH significantly lower than phys¬ iologic, they are sintered (heated) to 600°C to 700°C This results in a hard, strong, functionally nonresorbable ma¬ terial that has gained the widest clini¬ cal application primarily within the fields of dentistry and oral surgery.14"16
at
.
Preformed ceramic HA is available in two forms: dense and porous. Po¬ rous ceramic HA permits bone in¬ growth if the pores have a minimum diameter of 100 µ . Interconnecting pores of 40 to 100 µ are necessary for the development of functioning haversian systems.18,22,23 Certain species of coral in the genus Porites have regu¬ larly patterned 200-µ pores, with interconnecting pores of 50 to 60 µ . This structure proved to be ideal for osseous ingrowth, and commercial im¬ plants based on this coralline structure have been developed and marketed.23'25 Dense or porous ceramic HA is avail¬ able as blocks or granules. Blocks, used as substitutes for bone grafts in orthognathic surgery, are difficult to shape such that they accurately con¬ form to facial skeletal defects and are very brittle in the porous form. Gran¬ ules are difficult to confine within the reconstruction area and are not struc¬ turally stabile until ingrown by fibroosseous tissue. This ingrowth and sta¬ bilization can take several weeks to several months (Fig 2). Hydroxyapatite cement, which was evaluated in this study, is significantly different from the ceramic HA com¬ monly used in clinical practice. Hy¬ droxyapatite cement, which is noncer¬ amic, is produced by direct crys¬ tallization of HA in vivo, and it does not require heating for the formation of a structurally stable implant. The solubility of nonsintered HA is greater than that of ceramic HA, but sparingly soluble preparations are still possible.19 Prior to the development of HAC, nonceramic forms of HA had not been used clinically due to their bioresorbability. The material used in this study can be mixed with water, blood, saline, or weak phosphoric acid to create a "paste." The paste can then be used to fill facial-skeletal defects and can be contoured The intraoperatively. strength of HAC is less than that of ceramic HA, but, based on stress test¬ ing, it appears to be sufficient for the
Ca„(P04)2 + CaHP04
H20
-*·
0%( 4)3
Fig 3.—Chemical equation showing chemical reaction to form hydroxyapatite. Water is a vehi¬ cle; reactants dissolve and then reprecipitate as hydroxyapatite after setting in situ.
Fig 4.—The dry hydroxyapatite cement.
Fig 2. —Ceramic hydroxyapatite granules.
reconstruction of
bone.2"
non-stress-bearing
Chemistry and Properties of HAC
The major components of HAC are tetracalcium phosphate (TTCP) and dicalcium phosphate anhydrous or dicalcium phosphate dihydrate (DCP). These components react in an aqueous environment to form HA, and a simpli¬ fied chemical equation for this reaction is presented in Fig 3. The TTCP is a more basic salt than is HA, and the DCP is more acidic than is HA. To¬ gether, TTCP and DCP can form HA as the only product if the stoichiometry of the cement paste is in accordance with the chemical equation in Fig 3. Under in vitro conditions at 37°C, pure HAC (Figs 4 and 5) sets in ap¬ proximately 15 minutes, maintaining a pH during the setting phase in the range of 6.5 to 8.0.19 In a recent investigation,2" pow¬ der roentgenographic diffraction tech¬ niques were used to study the extent of the cement setting reaction as a function of time during the 24-hour period following mixing. The compres¬ sive strengths of the samples during the same period were also determined. The results showed that the cement sets within 15 minutes and the chemi¬ cal reaction was completed within 4 hours, at which time the only reaction product was HA with no by-products. A maximum compressive strength of
380
kg/cm2 (SD
=
13
kg/cm2;
=
8) (5400
pounds per square inch; 37 MPa was also developed at this time. Under scanning electron microscopic exami¬ nation, the set HAC is composed of small petal-like crystals, which result in a microporous structure. The bulk density of the set cement is approxi¬ mately 2.3, which corresponds to a pore content of 45% of the entire im¬ plant volume.19 Results from dye pene¬ tration tests and scanning electron mi¬
croscopic examinations*28 suggest that
the average effective pore diameter is in the order of 2 to 5 nm. This pore size will allow permeation of the set cement by ionic materials and dyes such as méthylène blue, but prevents the pas¬ sage of bacteria. MATERIALS AND METHODS
Experimental Design Thirty-six disks of HAC were fabricated by filling a 1-cm-diameter cylindrical mold with the material, compressing it to "firm¬ ness" (by hand) and allowing it to set in a moist environment for 12 hours. The solid cylinders of cement were then sectioned into disks 1.0 mm thick, using a saw. To evaluate the relationship of implant résorption to im¬ plant surface area, macropores were embed¬ ded within some of the implants by mixing the cement with sucrose granules prior to placement into the cylindrical molds. After setting, the sucrose granules were removed by dissolution in water. In this fashion, three sets of 12 disks were produced, containing 0%, 109f, and 20% macropore content by vol¬ ume, respectively. The macropores provided
Fig 5.—The with water.
hydroxyapatite cement after mixing
channels into which bone could grow and ef¬ fectively increased the surface area of the disks to evaluate the relationship of résorp¬ tion to surface area. Nine healthy adult mixed-breed cats served as experimental subjects. The proto¬ col and guidelines for this study were ap¬ proved by the Institutional Animal Care and Use Review Committee of Northwestern University Medical School, Chicago, 111. The animals were procured by the Northwestern University Medical School Center for Ex¬ perimental Animal Surgery. The preoper¬ ative and postoperative care of these animals was overseen by the University veterinari¬ ans to ensure proper and humane treatment. Surgery was performed under sterile condi¬ tions, using endotracheal halothane anesthe¬ sia after induction with intravenous 4% sodi¬ um pentobarbital (1 mL/2.25 kg). After operative procedures, the animals received
buprenorphine hydrochloride (Buprenex), 0.01 mg/kg intramuscularly every 12 hours over a period of 48 hours for pain control. A single preoperative dose of cephalothin sodi¬ um (Keflin), 40 mg/kg, was given intrave¬ nously for antibiotic coverage. Through a single 2-cm midline scalp inci¬ sion positioned between the ears of each ani¬ mal, disks composed of HAC were implanted
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
Fig 6.—A hydroxyapatite cement disk being planted intramuscularly.
im¬
subcutaneously, intramuscularly, just out¬ side the periosteum of the skull, and directly onto the surface of the calvarium below the
periosteum (Fig 6). The animals were sacri¬ ficed by lethal injection at 3, 6, and 9 months after implantation. Each group of animals had disks of all pore percentages represented in each tissue plane.
Fig 7.—A decalcified section of a disk implanted intramuscularly 6 months prior to killing of the animal. The disk has been dissolved by the décalcification process and has been present in the hydroxyapa¬ tite (HA) space. The inflammatory response is consistent with a postsurgical wound, and there is no evidence of foreign body giant cell formation (hematoxylin-eosin, original magnification 40).
Histologie Processing and Examination
Disks embedded within the soft tissue removed from the animals with a sur¬ rounding "cuff of tissue. The disks placed subperiosteally were firmly fixed to the un¬ derlying skull, and, therefore, the entire sec¬ tion of calvarium underlying the disks was cut free from the rest of the cranium, using a saw (Stryker). The disks and their surround¬ ing tissue were placed in a 10% buffered formaldehyde solution for fixation. To better evaluate inflammatory re¬ sponse, three disks implanted in muscle were decalcified by placing the disks in a 12% hy¬ drochloric acid solution until soft enough for sectioning. The decalcified specimen was then paraffin-embedded, sectioned, and stained with hematoxylin-eosin by standard technique.29'30 This process of décalcification removes most of the implant material. The remainder of the disks were not decalcified to preserve the HA implant for microscopic ex¬ amination. These specimens were embedded in methylmethacrylate, sectioned with a dia¬ mond-tipped blade, ground to a thickness of 40 µ , and stained with Paragon stain (Ladd Multiple Stain, Burlington, Vt), as described by Dunneisen."1 This method of staining al¬ lowed good differentiation between fully mineralized bone, osteoid and soft tissue. Examination of both the decalcified and undecalcified sections was carried out micro¬ scopically. The decalcified specimens were examined for the amount of inflammation and fibrous encapsulation of the implants. The remaining specimens were examined for disk résorption, osseointegration of the disks to
were
underlying calvarium, and implant re¬ placement by bone. Formal histometric anal¬ ysis of bone growth into the implants was not carried out. Radiograms were obtained on fine-grain film (Kodak SRE. Eastman Kodak Co, Rochester, NY) with a radiographie the
machine (Faxitron, Hewlett-Packard Co, McMinnville, Ore) set at 40 keV, 3 mA, and 120 seconds at 9-cm distance. Radiograms suitable for microscopic examination were obtained with the same Faxitron machine at 40 keV and 12-minute exposure and exam¬ 60 ined with a light microscope at
magnification.
postsurgical wound (Fig 7). There was no fibrous encapsulation of the im¬ plants, regardless of the tissue plane within which the disks were implant¬ ed, and there was no foreign body giant cell response in any of the specimens. The undecalcified soft-tissue speci¬ mens demonstrated progressive ré¬ sorption of the implants, with a loss of volume,
over
the 9 months of the
study. Specimens with the greater pore content demonstrated
RESULTS All animals survived the entire du¬ ration of the study. There were no toxic reactions, wound infections, im¬ plant extrusions, and no disks migrat¬ ed from the original sites of implanta¬ tion. Gross examination of the disks demonstrated no fibrous encapsula¬ tion. Those disks implanted onto the surface of the calvarium were firmly fused to the underlying skull. Decalcified histologie examination was carried out only on disks im¬ planted intramuscularly. The decalci¬ fied soft-tissue specimens revealed mild inflammatory response, consist¬ ing of polymorphonuclear leukocytes and monocytes. This inflammatory re¬ sponse resolved within 6 months after implantation and was consistent with a
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
greater
ré¬
sorption within each time interval. The absence of significant fibrous encapsu¬
lation was also confirmed on these sections. No evidence of osteogenesis was seen. In contrast to the implants placed in the soft tissues, the speci¬ mens implanted subperiosteally dem¬ onstrated numerous foci of new bone formation within the spaces between the disks and the calvarium (Fig 8). There was bone ingrowth into the disks over the surfaces in contact with the skull (Fig 9). Disks implanted sub¬ periosteally also demonstrated bone formation over their external surfaces, which was not in direct contact with the skull, but was covered by
periosteum. Roentgenographic examination of the disks implanted within the soft tissue confirmed progressive resorp-
Fig 8. —Tan lamellar bone of the skull (lower half of the figure) with a layer of recently deposited hypercellular bone on its surface 6 months after disk implantation
can be seen. The implant can be seen as the dark mottled material in the upper half of the figure. Bone has grown into the implant from the skull surface, and areas of osteoid (magenta) can be seen around the resorbing islands of hydroxyapatite cement (dark material) (Paragon stain,
original magnification 40).
tion in direct relation to greater pore content and time implanted. There was no roentgenographic evidence of bone deposition in any of the disks implant¬ ed within the soft tissue. However, for
subperiosteal disks, microroentgenograms of the junction between the
disks and the skull confirmed that the deposited material was calcified and probably represented bone (Fig 10). Attenuation of the disk surfaces in contact with calvarium without a loss of disk volume was seen, indicating that the implant had been replaced by incompletely calcified osteoid. The amount of calcified material within the disk-skull junction increased over time, as did the amount of disk re¬ placement by less radiodense material
(presumably bone). COMMENT
Hydroxyapatite represents a highly biocompatible synthetic implant mate¬
rial for skeletal reconstruction. This biocompatibility stems from the fact that calcium phosphate is found within bone in the form of HA, and it is, therefore, a chemically natural materi¬ al. Widespread application of HA to bone replacement has been limited by the poor ability of HA to withstand
Fig 9.— A junction between a 0% pore disk and skull 3 months after implantation. A layer of tan bone can be seen over the inferior surface of the dark implant (upper half of illustration). Bone can also be seen growing into the implant without a loss of implant volume (far left of illustration). Note the osteocytes within the layer of recently deposited bone on the undersurface of the implant (Paragon stain, original magnification 100).
shear
forces, which prevents its use in stress-bearing bones, and the need for préfabrication as a hard implant, which limits its contour adaptability. Since HAC can be sculpted intraoperatively, it potentially eliminates the need for préfabrication and may solve the con¬ tour adaptability problem faced by ce¬
ramic HA preparations.19'2" The histologie response to the HAC when implanted within soft tissue was consistent with the response seen with other forms of prefabricated, commer¬ cial HA. ''2""23 Lack of significant fibrous encapsulation, minimal inflammatory response, and an absence of foreign body giant cell formation would be expected of all ceramic or nonceramic HA implants. In this study, the inflam¬ matory response that had resolved by 6 months was consistent with a postsurgical wound. The absence of fibrous encapsulation and minimal inflamma¬ tion compare favorably with what is seen with polymer implants such as
methylmethacrylate or Proplast.8
It appears that HAC is more resorbable than ceramic HA. The in¬ creased résorption of disks with great¬ er pore content indicates that resorbability varies directly with surface area. There was no evidence of osteo-
Fig 10.—A microroentgenogram of the junction between a 10% pore hydroxyapatite disk (HA) and the skull (S) 6 months after implantation. There is a layer of radiodense new bone forma¬ tion on the surface of the skull. The arrows indi¬ cate calcified osseous tissue joining the disk to the calvarium (magnification 40).
genesis in any of the HAC disks, which
is consistent with data on ceramic HA implants. It would be expected that, when implanted solely within soft tis¬ sue, HAC would not be replaced by bone over time and would eventually be resorbed. Although no evidence of true osteo¬ genesis was found in any disk speci¬ mens, the HAC was found to be osteoconductive when implanted in subperi¬ osteal pockets or onto the surface of the calvarium. Bone grew onto the
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
surfaces of the disks and into the im¬
plant material
it was slowly re¬ sorbed. The external surfaces of the disks implanted onto the calvarium (beneath the periosteum, but not in direct contact with bone) also demon¬ strated foci of bone formation. The bone formation seen over the external surfaces of the subperiosteal implants does not represent osteogenesis but, rather, bone formation generated by the periosteum itself, which has a sig¬ nificant endogenous osteogenic poten¬ tial.32 Although the cement does not appear to form a permanent implant if surrounded only by soft tissue, its osteoconductive and resorptive prop¬ erties suggest that it has the potential to function as a successful implant when in direct contact with viable bone. When in contact with viable bone, the cement is replaced, at least as
in part, with osseous tissue, without a significant loss of volume. This proper¬ ty is unique when compared with com¬ mercial ceramic HA preparations. Several conclusions can be drawn from this initial study. First, the fact that this material is a cement and can be sculptured in vivo is a significant advantage over preformed ceramic HA blocks or granules, when facial contour reconstruction or augmentation is con¬ templated. Second, when HAC is placed into contact with viable bone or periosteum, it is replaced by bone as the implant is resorbed. While HAC probably does not have sufficient shear-strength resistance to function in the reconstruction of stress-bearing bones, the cement does appear to be sufficiently structurally stable for re¬ construction and augmentation of nonstress-bearing portions of the craniofa-
cial skeleton.
Finally,
the
inflam¬
matory response to this cement is less
than to that seen with polymer im¬ plants, and it is comparable to that seen with clinically used forms of ce¬ ramic HA. Based on these favorable initial findings, HAC was evaluated experimentally in part II24 for frontal sinus obliteration and frontal bone reconstruction. The research presented in this study was not initiated or supported by any corporation. Partial funding was provided by the Yul Brynner Re¬ search Foundation. Inc. Chicago, 111, with techni¬ cal and material support from the American Den¬ tal Association Health Foundation Paffenbarger Research Center at the National Institute of Standards and Technology, Gaithersburg, Md, under funding from the National Institutes of Health (Bethesda, Md) grant DE-05030. Hydroxyapatite cement is a proprietary materi¬ al of the American Dental Association Health Foundation. At present, the cement is only avail¬ able for research purposes.
References 1. Fallopius G. Opera omnia Francofurti. Wecheli A, ed. 1600;1. 2. Gage EL. Vitallium cranioplasty. WV Med J.
1971;67:325.
3. Gordon DS, Blair GAS. Titanium cranioplasty. BMJ. 1974;2:478. 4. Janeke JB, Komorn RM, Cohn AM. Proplast in cavity obliteration and soft tissue augmentation. Arch Otolaryngol. 1974;100:24-27. 5. Wolfe SA. Correction of persistent lower eyelid deformity caused by a displaced orbital floor implant. Ann Plast Surg. 1979;2:448. 6. Wolfe SA. Correction of lower eyelid deformity caused by multiple extrusions of alloplastic orbital floor implants. Plast Reconstr Surg. 1981;68:429. 7. Kent JN, Zide MF. Wound healing: bone and biomaterials. Otolaryngol Clin North Am. 1984;
17:273-319. 8. Rish BL, Dillon JD, Meirowsky AM, et al. Cranioplasty: a review of 1030 cases of penetrating head injury. Neurosurgery. 1979;4:381-390. 9. Jackson IT, Helden G, Marx R. Skull bone grafts in maxillofacial and craniofacial surgery. J Oral Maxillofac Surg. 1986;44:949-955. 10. Oklund SA, Prolo DJ, Gutierrez RV, King SE. Quantitative comparisons of healing in cranial fresh autografts, frozen autografts, and processed autografts, and allografts in canine skull defects. Clin Orthoped. 1986;205:269-291. 11. Wolfe SA. Autogenous bone grafts versus alloplastic material in maxillofacial surgery. Clin Plast Surg. 1982;9:539-540. 12. Kent JN. Reconstruction of the alveolar ridge with hydroxyapatite. Dent Clin North Am.
1986;30:231-257. 13. Block MS, Kent JN, Ardoin RC, Davenport W. Mandibular augmentation in dogs with hy-
droxyapatite combined with demineralized bone. J
Oral Maxillofac Surg. 1987;45:414-420. 14. Jarcho M, Kay JF, Gumaer KI, et al. Tissue, cellular and subcellular events at a bone-ceramic hydroxyapatite interface. J Bioeng. 1977;1:79-92. 15. Jarcho M. Calcium phosphate ceramics as hard tissue prosthetics. Clin Orthop. 1981;157: 259-278. 16. Jarcho
et al. and characterization in form. J Mater Sci. 1976;
M, Bolen CH, Thomas MB,
Hydroxyapatite synthesis
dense polycrystalline 11:2027-2035. 17. Kent JN, Zide MF, Kay JF. Hydroxyapatite blocks and particles as bone graft substitutes in orthognathic and reconstructive surgery. J Oral
Maxillofac Surg. 1986;44:597-605. 18. Holmes RE. Bone regeneration within a coralline hydroxyapatite implant. Plast Reconstr Surg. 1979;63:626-633. 19. Brown WE, Chow LC. A new calcium phosphate, water-setting cement. In: Brown PW, ed. Cements Research Progress. Westerville, Ohio: American Ceramic Society; 1986:352-379. 20. Friedman CD, Costantino PD, Jones K, Chow LC, Pelzer HJ, Sisson GA. Hydroxyapatite cement, II: Obliteration and reconstruction of the cat frontal sinus. Arch Otolaryngol Head Neck
Surg. 1991;117:385-389. 21. Boyne JP, Fremming BD, Walsh R, Jarcho M. Evaluation of ceramic hydroxyapatite in femo-
ral defects. J Dent Res. 1978;57A:108. Abstract. 22. Ducheyne P, Hench L, Kagan A, et al. Effect of hydroxyapatite impregnation on skeletal bonding of porous coated implants. J Biomed Mater Res.
1980;14:225-237.
23. Holmes RE, Hagler HK. Porous hydroxyapatite as a bone graft substitute in cranial reconstruction: a histometric study. Plast Reconstr
Surg. 1988;81:662-671.
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
24. Grenga TE, Zins JE, Bauer TW. The rate of vascularization of coralline hydroxyapatite. Plast Reconstr Surg. 1989;84:245-249. 25. Piecuch JF, Fedorka NJ. Results of soft\x=req-\ tissue surgery over implanted replamineform hydroxyapatite. J Oral Maxillofac Surg. 1983;41: 801-806. 26. Fukase Y, Takagi S, Eanes ED, et al. Setting reactions and compressive strength of calcium phosphate cement. J Dent Res. 1989;68:398. Abstract. 27. Grunninger SW, Siew C, Chow LC, et al. Evaluation of the biocompatibility of a new calcium phosphate setting cement. J Dent Res. 1984;63:200. Abstract. 28. Sugawara A, Antonucci JM, Takagi S. Formation of hydroxyapatite in hydrogels from tetracalcium phosphate/dicalcium phosphate mixtures. J Dent Res. 1988;67:383. Abstract. 29. Sheehan DC, Hrapchak BB. Bone. In: Sheehan DC, Hrapchak BB, eds. Theory and Practice of Histochemistry. 2nd ed. St Louis, Mo: CV Mosby
Co; 1980;6:89-117. 30. Schenk RK, Olah AJ, Herrmann W. Preparation of calcified tissues for light microscopy. In: Dickson GR, ed. Methods of Calcified Tissue Preparation. Amsterdam, the Netherlands: Elsevier Science Publishers; 1984:1-56.
31. Dunneisen E. An improved technique for methylmethacrylate embedding and toluidine blue/ basic fuchsin staining of undecalcified bone. Presented at 33rd Annual Meeting of Orthopaedic Research Society, January 19-22,1987; San Francisco,
Calif. Abstract. 32. Wlodarski KH. Normal and heterotopic periosteum. Clin Orthop. 1989;241:265-277.
Cement
Chemistry and Histologic Properties
Peter D. Costantino, MD; Craig D. Friedman, MD; Kent Jones, MD; Lawrence C. Chow, PhD; Harold J. Pelzer, DDS, MD; George A. Sisson, Sr, MD
Hydroxyapatite cement is a unique calphosphate preparation that can be shaped intraoperatively and sets in vivo to an implant composed of microporous hydroxyapatite. The histologic response to this cement was evaluated by implanting \s=b\
cium
disks made of this material within the heads of nine cats. Three sets of 12 hydroxyapatite cement disks were produced containing 0%, 10%, and 20% macropores by volume, respectively. The disks were implanted subcutaneously, intramuscularly, above the periosteum of the skull, and directly onto the surface of the calvarium. Each macropore percentage was represented in each tissue plane, and animals were killed up to 9 months postoperatively. There were no toxic reactions, implants extruded, or wound infections. Histologic examination of the implant\p=m-\soft-tissue interfaces revealed a
Accepted for publication November 2,1990. From the Department of Otolaryngology\p=m-\Head and Neck Surgery, Northwestern University Medical School, Chicago, Ill (Drs Costantino, Jones, Pelzer, and Sisson); Section of Otolaryngology\p=m-\ Head and Neck Surgery, Yale University Medical School, New Haven, Conn (Dr Friedman); and the
American Dental Association Health Foundation, Paffenbarger Research Institute at the National Institute of Standards and Technology, Gaithersburg, Md (Dr Chow). Presented at the Fifth International Symposium on Facial Plastic Reconstructive Surgery of the Head and Neck, June 27,1989, Toronto, Ontario. Reprint requests to Section of Otolaryngology\p=m-\ Head and Neck Surgery, Yale University School of 333 Cedar St, New Haven, Medicine, PO Box CT 06510 (Dr Friedman).
3333,
inflammatory response without foreign body reaction. The disks were resorbed over time in direct proportion to their macropore content (surface areas) in all groups except for those disks placed directly onto the surface of the calvarium below the periosteum. In this group, numerous foci of bone formed at the skull-implant interface, with variable replacement of the deep surface of these implants by bone. Implant replacement by bone is postulated to occur through a combination of implant resorption coupled with osteoconduction. Based on these properties, hydroxyapatite cement may prove useful when applied to the reconstruction of non\p=m-\stress-bearing transient
skeletal tissue.
(Arch Otolaryngol Head 117:379-384)
Neck
Surg. 1991;
Alloplastic implants
have been used craniofacial skeleton since 1600, when Fallopius to reconstruct the
implanted
a
gold plate
to
repair
a
cranial defect.1 Since that time, con¬ tour reconstruction of the non-stressbearing, craniofacial skeleton has con¬ tinued to be a technical and materials problem. Metals are difficult to shape and are hampered by problems such as infection and corrosion.23 Polymers such as silicone, Proplast, or methyl-
methacrylate are encapsulated by scar resulting in significant rates of implant infection and/or extrusion.4"* Biologic materials, such as autogenous bone grafts, may cause donor site mor¬ bidity, suffer from significant postim¬ plantation résorption, and are difficult to accurately conform to skeletal de¬
tissue
fects.9"11 At present, there is no ideal natural or synthetic material for con¬ tour reconstruction of the craniofacial
skeleton. Of those alloplastic materials used to augment and replace the facial skele¬ ton, the most promising and well tol¬ erated are calcium phosphate-based compounds.'12"14 The majority of the human skeleton is composed of calcium phosphate in the form of hydroxyapa¬ tite (HA), and processing techniques developed during the early 1970s have provided HA preparations with suffi¬ cient structural strength to serve as bone substitutes.1" Since the mid1970s, various preformed HA prepara¬ tions have been used in a variety of clinical applications within medicine and dentistry.1'1"18 The applicability of these HA preparations, however, has been limited, because these implants had to be preformed as a hard materi¬
al.18
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
Fig 1. Solubility phase diagram for various cal¬ cium phosphates at 37°C. OHAp indicates hydroxyapatite; ß-TCP, beta-tricalcium phos¬ phate; TTCP, tetracalcium phosphate; OCP, octacalcium phosphate; DDCP, dihydrate dicalcium phosphate; and ADCP, anhydrous dicalcium phosphate. —
Recently, a new type of calcium phosphate cement that sets to HA in
vivo has been developed19 at the Paffenbarger Research Center of the American Dental Association Health Foundation (ADAHF) located at the National Institutes of Standards and Technology, Gaithersburg, Md. This material is referred to as HA cement (HAC) and is particularly unique in that it can be applied intraoperatively
paste and, subsequently, sets to a structurally stabile implant composed of microporous HA. All other forms of HA, to our knowledge, must be pre¬ formed as a hard material prior to implantation. The purpose of this study was to test the histologie response to HAC in a large animal model. Application of as a
HAC to frontal sinus obliteration and reconstruction has also been per¬ formed and is described in part II20 of this study. The hypotheses tested herein were that HAC would be highly tissue compatible, would not be resorbable, and would demonstrate properties potentially suitable for use in the augmentation of nonstress-bearing areas of the craniofacial skeleton. BACKGROUND Biologic Characteristics of HA Compounds
The HA is
general chemical composition of [Cairj(P04)6(OH)2], and it essen¬
tially represents interlinked calcium phosphate molecules. Hydroxyapatite
is the primary mineral component of teeth and bone and exists in bone in equilibrium with other calcium phos¬ phate salts. At a pH of 7.4, it is the least soluble of the naturally occurring calcium phosphate salts (Fig 1). Its stable macromolecular structure, along with its low solubility at physiologic pH,1'19 results in a relative resistance to résorption in vivo, depending on the method of preparation. All forms of HA are noteworthy for their excellent biocompatibility. Hydroxyapatite does not cause a foreign body giant cell reaction, a sustained inflammatory response, toxic reac¬ tions, or an increase in serum calcium or phosphate levels. The most impor¬ tant characteristic of HA compounds is their interaction with osseous tissue. Some studies have claimed that the presence of HA compounds within the body induces osteogenesis (de novo formation of new bone), but definitive studies21 using bone marrow and HA in implanted Millipore filters showed no osteogenic effect by HA in the absence of bone marrow cells.21 Although HA is not osteogenic, it is osteoconductive, in that it can serve as a scaffold on which bone can grow. If macropores (200 to 300 µ ) are fabri¬ cated into an HA implant, bone grows into the macropores. "',22"24 When im¬ planted into the cortex of long bones in dogs, preformed porous ceramic HA blocks becomes ingrown with osteoid tissue.1" Implanted blocks of HA can show a faster rate of bone ingrowth than can bone grafts,23 and similar studies have also shown that cartilage is able to grow into porous HA im¬ plants. Despite a lack of osteogenesis, a direct chemical bond forms between the implant and bone without interven¬ ing fibrous tissue. This bonding of im¬ plant to bone is referred to as
osseointegration.18'22
Types of HA:
Ceramic and Nonceramic
Hydroxyapatite can be fabricated in ceramic and nonceramic forms. Since the mid-1970s, HA has been used clini¬ cally only as a ceramic preparation. The ceramic forms of HA are heated to fuse individual HA crystals to each other through a process called sinter¬ ing. After HA crystals are fabricated
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
a pH significantly lower than phys¬ iologic, they are sintered (heated) to 600°C to 700°C This results in a hard, strong, functionally nonresorbable ma¬ terial that has gained the widest clini¬ cal application primarily within the fields of dentistry and oral surgery.14"16
at
.
Preformed ceramic HA is available in two forms: dense and porous. Po¬ rous ceramic HA permits bone in¬ growth if the pores have a minimum diameter of 100 µ . Interconnecting pores of 40 to 100 µ are necessary for the development of functioning haversian systems.18,22,23 Certain species of coral in the genus Porites have regu¬ larly patterned 200-µ pores, with interconnecting pores of 50 to 60 µ . This structure proved to be ideal for osseous ingrowth, and commercial im¬ plants based on this coralline structure have been developed and marketed.23'25 Dense or porous ceramic HA is avail¬ able as blocks or granules. Blocks, used as substitutes for bone grafts in orthognathic surgery, are difficult to shape such that they accurately con¬ form to facial skeletal defects and are very brittle in the porous form. Gran¬ ules are difficult to confine within the reconstruction area and are not struc¬ turally stabile until ingrown by fibroosseous tissue. This ingrowth and sta¬ bilization can take several weeks to several months (Fig 2). Hydroxyapatite cement, which was evaluated in this study, is significantly different from the ceramic HA com¬ monly used in clinical practice. Hy¬ droxyapatite cement, which is noncer¬ amic, is produced by direct crys¬ tallization of HA in vivo, and it does not require heating for the formation of a structurally stable implant. The solubility of nonsintered HA is greater than that of ceramic HA, but sparingly soluble preparations are still possible.19 Prior to the development of HAC, nonceramic forms of HA had not been used clinically due to their bioresorbability. The material used in this study can be mixed with water, blood, saline, or weak phosphoric acid to create a "paste." The paste can then be used to fill facial-skeletal defects and can be contoured The intraoperatively. strength of HAC is less than that of ceramic HA, but, based on stress test¬ ing, it appears to be sufficient for the
Ca„(P04)2 + CaHP04
H20
-*·
0%( 4)3
Fig 3.—Chemical equation showing chemical reaction to form hydroxyapatite. Water is a vehi¬ cle; reactants dissolve and then reprecipitate as hydroxyapatite after setting in situ.
Fig 4.—The dry hydroxyapatite cement.
Fig 2. —Ceramic hydroxyapatite granules.
reconstruction of
bone.2"
non-stress-bearing
Chemistry and Properties of HAC
The major components of HAC are tetracalcium phosphate (TTCP) and dicalcium phosphate anhydrous or dicalcium phosphate dihydrate (DCP). These components react in an aqueous environment to form HA, and a simpli¬ fied chemical equation for this reaction is presented in Fig 3. The TTCP is a more basic salt than is HA, and the DCP is more acidic than is HA. To¬ gether, TTCP and DCP can form HA as the only product if the stoichiometry of the cement paste is in accordance with the chemical equation in Fig 3. Under in vitro conditions at 37°C, pure HAC (Figs 4 and 5) sets in ap¬ proximately 15 minutes, maintaining a pH during the setting phase in the range of 6.5 to 8.0.19 In a recent investigation,2" pow¬ der roentgenographic diffraction tech¬ niques were used to study the extent of the cement setting reaction as a function of time during the 24-hour period following mixing. The compres¬ sive strengths of the samples during the same period were also determined. The results showed that the cement sets within 15 minutes and the chemi¬ cal reaction was completed within 4 hours, at which time the only reaction product was HA with no by-products. A maximum compressive strength of
380
kg/cm2 (SD
=
13
kg/cm2;
=
8) (5400
pounds per square inch; 37 MPa was also developed at this time. Under scanning electron microscopic exami¬ nation, the set HAC is composed of small petal-like crystals, which result in a microporous structure. The bulk density of the set cement is approxi¬ mately 2.3, which corresponds to a pore content of 45% of the entire im¬ plant volume.19 Results from dye pene¬ tration tests and scanning electron mi¬
croscopic examinations*28 suggest that
the average effective pore diameter is in the order of 2 to 5 nm. This pore size will allow permeation of the set cement by ionic materials and dyes such as méthylène blue, but prevents the pas¬ sage of bacteria. MATERIALS AND METHODS
Experimental Design Thirty-six disks of HAC were fabricated by filling a 1-cm-diameter cylindrical mold with the material, compressing it to "firm¬ ness" (by hand) and allowing it to set in a moist environment for 12 hours. The solid cylinders of cement were then sectioned into disks 1.0 mm thick, using a saw. To evaluate the relationship of implant résorption to im¬ plant surface area, macropores were embed¬ ded within some of the implants by mixing the cement with sucrose granules prior to placement into the cylindrical molds. After setting, the sucrose granules were removed by dissolution in water. In this fashion, three sets of 12 disks were produced, containing 0%, 109f, and 20% macropore content by vol¬ ume, respectively. The macropores provided
Fig 5.—The with water.
hydroxyapatite cement after mixing
channels into which bone could grow and ef¬ fectively increased the surface area of the disks to evaluate the relationship of résorp¬ tion to surface area. Nine healthy adult mixed-breed cats served as experimental subjects. The proto¬ col and guidelines for this study were ap¬ proved by the Institutional Animal Care and Use Review Committee of Northwestern University Medical School, Chicago, 111. The animals were procured by the Northwestern University Medical School Center for Ex¬ perimental Animal Surgery. The preoper¬ ative and postoperative care of these animals was overseen by the University veterinari¬ ans to ensure proper and humane treatment. Surgery was performed under sterile condi¬ tions, using endotracheal halothane anesthe¬ sia after induction with intravenous 4% sodi¬ um pentobarbital (1 mL/2.25 kg). After operative procedures, the animals received
buprenorphine hydrochloride (Buprenex), 0.01 mg/kg intramuscularly every 12 hours over a period of 48 hours for pain control. A single preoperative dose of cephalothin sodi¬ um (Keflin), 40 mg/kg, was given intrave¬ nously for antibiotic coverage. Through a single 2-cm midline scalp inci¬ sion positioned between the ears of each ani¬ mal, disks composed of HAC were implanted
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
Fig 6.—A hydroxyapatite cement disk being planted intramuscularly.
im¬
subcutaneously, intramuscularly, just out¬ side the periosteum of the skull, and directly onto the surface of the calvarium below the
periosteum (Fig 6). The animals were sacri¬ ficed by lethal injection at 3, 6, and 9 months after implantation. Each group of animals had disks of all pore percentages represented in each tissue plane.
Fig 7.—A decalcified section of a disk implanted intramuscularly 6 months prior to killing of the animal. The disk has been dissolved by the décalcification process and has been present in the hydroxyapa¬ tite (HA) space. The inflammatory response is consistent with a postsurgical wound, and there is no evidence of foreign body giant cell formation (hematoxylin-eosin, original magnification 40).
Histologie Processing and Examination
Disks embedded within the soft tissue removed from the animals with a sur¬ rounding "cuff of tissue. The disks placed subperiosteally were firmly fixed to the un¬ derlying skull, and, therefore, the entire sec¬ tion of calvarium underlying the disks was cut free from the rest of the cranium, using a saw (Stryker). The disks and their surround¬ ing tissue were placed in a 10% buffered formaldehyde solution for fixation. To better evaluate inflammatory re¬ sponse, three disks implanted in muscle were decalcified by placing the disks in a 12% hy¬ drochloric acid solution until soft enough for sectioning. The decalcified specimen was then paraffin-embedded, sectioned, and stained with hematoxylin-eosin by standard technique.29'30 This process of décalcification removes most of the implant material. The remainder of the disks were not decalcified to preserve the HA implant for microscopic ex¬ amination. These specimens were embedded in methylmethacrylate, sectioned with a dia¬ mond-tipped blade, ground to a thickness of 40 µ , and stained with Paragon stain (Ladd Multiple Stain, Burlington, Vt), as described by Dunneisen."1 This method of staining al¬ lowed good differentiation between fully mineralized bone, osteoid and soft tissue. Examination of both the decalcified and undecalcified sections was carried out micro¬ scopically. The decalcified specimens were examined for the amount of inflammation and fibrous encapsulation of the implants. The remaining specimens were examined for disk résorption, osseointegration of the disks to
were
underlying calvarium, and implant re¬ placement by bone. Formal histometric anal¬ ysis of bone growth into the implants was not carried out. Radiograms were obtained on fine-grain film (Kodak SRE. Eastman Kodak Co, Rochester, NY) with a radiographie the
machine (Faxitron, Hewlett-Packard Co, McMinnville, Ore) set at 40 keV, 3 mA, and 120 seconds at 9-cm distance. Radiograms suitable for microscopic examination were obtained with the same Faxitron machine at 40 keV and 12-minute exposure and exam¬ 60 ined with a light microscope at
magnification.
postsurgical wound (Fig 7). There was no fibrous encapsulation of the im¬ plants, regardless of the tissue plane within which the disks were implant¬ ed, and there was no foreign body giant cell response in any of the specimens. The undecalcified soft-tissue speci¬ mens demonstrated progressive ré¬ sorption of the implants, with a loss of volume,
over
the 9 months of the
study. Specimens with the greater pore content demonstrated
RESULTS All animals survived the entire du¬ ration of the study. There were no toxic reactions, wound infections, im¬ plant extrusions, and no disks migrat¬ ed from the original sites of implanta¬ tion. Gross examination of the disks demonstrated no fibrous encapsula¬ tion. Those disks implanted onto the surface of the calvarium were firmly fused to the underlying skull. Decalcified histologie examination was carried out only on disks im¬ planted intramuscularly. The decalci¬ fied soft-tissue specimens revealed mild inflammatory response, consist¬ ing of polymorphonuclear leukocytes and monocytes. This inflammatory re¬ sponse resolved within 6 months after implantation and was consistent with a
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
greater
ré¬
sorption within each time interval. The absence of significant fibrous encapsu¬
lation was also confirmed on these sections. No evidence of osteogenesis was seen. In contrast to the implants placed in the soft tissues, the speci¬ mens implanted subperiosteally dem¬ onstrated numerous foci of new bone formation within the spaces between the disks and the calvarium (Fig 8). There was bone ingrowth into the disks over the surfaces in contact with the skull (Fig 9). Disks implanted sub¬ periosteally also demonstrated bone formation over their external surfaces, which was not in direct contact with the skull, but was covered by
periosteum. Roentgenographic examination of the disks implanted within the soft tissue confirmed progressive resorp-
Fig 8. —Tan lamellar bone of the skull (lower half of the figure) with a layer of recently deposited hypercellular bone on its surface 6 months after disk implantation
can be seen. The implant can be seen as the dark mottled material in the upper half of the figure. Bone has grown into the implant from the skull surface, and areas of osteoid (magenta) can be seen around the resorbing islands of hydroxyapatite cement (dark material) (Paragon stain,
original magnification 40).
tion in direct relation to greater pore content and time implanted. There was no roentgenographic evidence of bone deposition in any of the disks implant¬ ed within the soft tissue. However, for
subperiosteal disks, microroentgenograms of the junction between the
disks and the skull confirmed that the deposited material was calcified and probably represented bone (Fig 10). Attenuation of the disk surfaces in contact with calvarium without a loss of disk volume was seen, indicating that the implant had been replaced by incompletely calcified osteoid. The amount of calcified material within the disk-skull junction increased over time, as did the amount of disk re¬ placement by less radiodense material
(presumably bone). COMMENT
Hydroxyapatite represents a highly biocompatible synthetic implant mate¬
rial for skeletal reconstruction. This biocompatibility stems from the fact that calcium phosphate is found within bone in the form of HA, and it is, therefore, a chemically natural materi¬ al. Widespread application of HA to bone replacement has been limited by the poor ability of HA to withstand
Fig 9.— A junction between a 0% pore disk and skull 3 months after implantation. A layer of tan bone can be seen over the inferior surface of the dark implant (upper half of illustration). Bone can also be seen growing into the implant without a loss of implant volume (far left of illustration). Note the osteocytes within the layer of recently deposited bone on the undersurface of the implant (Paragon stain, original magnification 100).
shear
forces, which prevents its use in stress-bearing bones, and the need for préfabrication as a hard implant, which limits its contour adaptability. Since HAC can be sculpted intraoperatively, it potentially eliminates the need for préfabrication and may solve the con¬ tour adaptability problem faced by ce¬
ramic HA preparations.19'2" The histologie response to the HAC when implanted within soft tissue was consistent with the response seen with other forms of prefabricated, commer¬ cial HA. ''2""23 Lack of significant fibrous encapsulation, minimal inflammatory response, and an absence of foreign body giant cell formation would be expected of all ceramic or nonceramic HA implants. In this study, the inflam¬ matory response that had resolved by 6 months was consistent with a postsurgical wound. The absence of fibrous encapsulation and minimal inflamma¬ tion compare favorably with what is seen with polymer implants such as
methylmethacrylate or Proplast.8
It appears that HAC is more resorbable than ceramic HA. The in¬ creased résorption of disks with great¬ er pore content indicates that resorbability varies directly with surface area. There was no evidence of osteo-
Fig 10.—A microroentgenogram of the junction between a 10% pore hydroxyapatite disk (HA) and the skull (S) 6 months after implantation. There is a layer of radiodense new bone forma¬ tion on the surface of the skull. The arrows indi¬ cate calcified osseous tissue joining the disk to the calvarium (magnification 40).
genesis in any of the HAC disks, which
is consistent with data on ceramic HA implants. It would be expected that, when implanted solely within soft tis¬ sue, HAC would not be replaced by bone over time and would eventually be resorbed. Although no evidence of true osteo¬ genesis was found in any disk speci¬ mens, the HAC was found to be osteoconductive when implanted in subperi¬ osteal pockets or onto the surface of the calvarium. Bone grew onto the
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
surfaces of the disks and into the im¬
plant material
it was slowly re¬ sorbed. The external surfaces of the disks implanted onto the calvarium (beneath the periosteum, but not in direct contact with bone) also demon¬ strated foci of bone formation. The bone formation seen over the external surfaces of the subperiosteal implants does not represent osteogenesis but, rather, bone formation generated by the periosteum itself, which has a sig¬ nificant endogenous osteogenic poten¬ tial.32 Although the cement does not appear to form a permanent implant if surrounded only by soft tissue, its osteoconductive and resorptive prop¬ erties suggest that it has the potential to function as a successful implant when in direct contact with viable bone. When in contact with viable bone, the cement is replaced, at least as
in part, with osseous tissue, without a significant loss of volume. This proper¬ ty is unique when compared with com¬ mercial ceramic HA preparations. Several conclusions can be drawn from this initial study. First, the fact that this material is a cement and can be sculptured in vivo is a significant advantage over preformed ceramic HA blocks or granules, when facial contour reconstruction or augmentation is con¬ templated. Second, when HAC is placed into contact with viable bone or periosteum, it is replaced by bone as the implant is resorbed. While HAC probably does not have sufficient shear-strength resistance to function in the reconstruction of stress-bearing bones, the cement does appear to be sufficiently structurally stable for re¬ construction and augmentation of nonstress-bearing portions of the craniofa-
cial skeleton.
Finally,
the
inflam¬
matory response to this cement is less
than to that seen with polymer im¬ plants, and it is comparable to that seen with clinically used forms of ce¬ ramic HA. Based on these favorable initial findings, HAC was evaluated experimentally in part II24 for frontal sinus obliteration and frontal bone reconstruction. The research presented in this study was not initiated or supported by any corporation. Partial funding was provided by the Yul Brynner Re¬ search Foundation. Inc. Chicago, 111, with techni¬ cal and material support from the American Den¬ tal Association Health Foundation Paffenbarger Research Center at the National Institute of Standards and Technology, Gaithersburg, Md, under funding from the National Institutes of Health (Bethesda, Md) grant DE-05030. Hydroxyapatite cement is a proprietary materi¬ al of the American Dental Association Health Foundation. At present, the cement is only avail¬ able for research purposes.
References 1. Fallopius G. Opera omnia Francofurti. Wecheli A, ed. 1600;1. 2. Gage EL. Vitallium cranioplasty. WV Med J.
1971;67:325.
3. Gordon DS, Blair GAS. Titanium cranioplasty. BMJ. 1974;2:478. 4. Janeke JB, Komorn RM, Cohn AM. Proplast in cavity obliteration and soft tissue augmentation. Arch Otolaryngol. 1974;100:24-27. 5. Wolfe SA. Correction of persistent lower eyelid deformity caused by a displaced orbital floor implant. Ann Plast Surg. 1979;2:448. 6. Wolfe SA. Correction of lower eyelid deformity caused by multiple extrusions of alloplastic orbital floor implants. Plast Reconstr Surg. 1981;68:429. 7. Kent JN, Zide MF. Wound healing: bone and biomaterials. Otolaryngol Clin North Am. 1984;
17:273-319. 8. Rish BL, Dillon JD, Meirowsky AM, et al. Cranioplasty: a review of 1030 cases of penetrating head injury. Neurosurgery. 1979;4:381-390. 9. Jackson IT, Helden G, Marx R. Skull bone grafts in maxillofacial and craniofacial surgery. J Oral Maxillofac Surg. 1986;44:949-955. 10. Oklund SA, Prolo DJ, Gutierrez RV, King SE. Quantitative comparisons of healing in cranial fresh autografts, frozen autografts, and processed autografts, and allografts in canine skull defects. Clin Orthoped. 1986;205:269-291. 11. Wolfe SA. Autogenous bone grafts versus alloplastic material in maxillofacial surgery. Clin Plast Surg. 1982;9:539-540. 12. Kent JN. Reconstruction of the alveolar ridge with hydroxyapatite. Dent Clin North Am.
1986;30:231-257. 13. Block MS, Kent JN, Ardoin RC, Davenport W. Mandibular augmentation in dogs with hy-
droxyapatite combined with demineralized bone. J
Oral Maxillofac Surg. 1987;45:414-420. 14. Jarcho M, Kay JF, Gumaer KI, et al. Tissue, cellular and subcellular events at a bone-ceramic hydroxyapatite interface. J Bioeng. 1977;1:79-92. 15. Jarcho M. Calcium phosphate ceramics as hard tissue prosthetics. Clin Orthop. 1981;157: 259-278. 16. Jarcho
et al. and characterization in form. J Mater Sci. 1976;
M, Bolen CH, Thomas MB,
Hydroxyapatite synthesis
dense polycrystalline 11:2027-2035. 17. Kent JN, Zide MF, Kay JF. Hydroxyapatite blocks and particles as bone graft substitutes in orthognathic and reconstructive surgery. J Oral
Maxillofac Surg. 1986;44:597-605. 18. Holmes RE. Bone regeneration within a coralline hydroxyapatite implant. Plast Reconstr Surg. 1979;63:626-633. 19. Brown WE, Chow LC. A new calcium phosphate, water-setting cement. In: Brown PW, ed. Cements Research Progress. Westerville, Ohio: American Ceramic Society; 1986:352-379. 20. Friedman CD, Costantino PD, Jones K, Chow LC, Pelzer HJ, Sisson GA. Hydroxyapatite cement, II: Obliteration and reconstruction of the cat frontal sinus. Arch Otolaryngol Head Neck
Surg. 1991;117:385-389. 21. Boyne JP, Fremming BD, Walsh R, Jarcho M. Evaluation of ceramic hydroxyapatite in femo-
ral defects. J Dent Res. 1978;57A:108. Abstract. 22. Ducheyne P, Hench L, Kagan A, et al. Effect of hydroxyapatite impregnation on skeletal bonding of porous coated implants. J Biomed Mater Res.
1980;14:225-237.
23. Holmes RE, Hagler HK. Porous hydroxyapatite as a bone graft substitute in cranial reconstruction: a histometric study. Plast Reconstr
Surg. 1988;81:662-671.
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/10/2015
24. Grenga TE, Zins JE, Bauer TW. The rate of vascularization of coralline hydroxyapatite. Plast Reconstr Surg. 1989;84:245-249. 25. Piecuch JF, Fedorka NJ. Results of soft\x=req-\ tissue surgery over implanted replamineform hydroxyapatite. J Oral Maxillofac Surg. 1983;41: 801-806. 26. Fukase Y, Takagi S, Eanes ED, et al. Setting reactions and compressive strength of calcium phosphate cement. J Dent Res. 1989;68:398. Abstract. 27. Grunninger SW, Siew C, Chow LC, et al. Evaluation of the biocompatibility of a new calcium phosphate setting cement. J Dent Res. 1984;63:200. Abstract. 28. Sugawara A, Antonucci JM, Takagi S. Formation of hydroxyapatite in hydrogels from tetracalcium phosphate/dicalcium phosphate mixtures. J Dent Res. 1988;67:383. Abstract. 29. Sheehan DC, Hrapchak BB. Bone. In: Sheehan DC, Hrapchak BB, eds. Theory and Practice of Histochemistry. 2nd ed. St Louis, Mo: CV Mosby
Co; 1980;6:89-117. 30. Schenk RK, Olah AJ, Herrmann W. Preparation of calcified tissues for light microscopy. In: Dickson GR, ed. Methods of Calcified Tissue Preparation. Amsterdam, the Netherlands: Elsevier Science Publishers; 1984:1-56.
31. Dunneisen E. An improved technique for methylmethacrylate embedding and toluidine blue/ basic fuchsin staining of undecalcified bone. Presented at 33rd Annual Meeting of Orthopaedic Research Society, January 19-22,1987; San Francisco,
Calif. Abstract. 32. Wlodarski KH. Normal and heterotopic periosteum. Clin Orthop. 1989;241:265-277.
