THE ANATOMICAL RECORD 245:410-425 (1996) Soft and Hard Tissue Response to Endosseous Dental Implants MAX A. LISTGARTEN University of Pennsylvania, School of Dental Medicine, Philadelphia, Pennsylvania ABSTRACT The last two decades have seen a remarkable growth in the development of dental implants and their incorporation into the practice of dentistry. This turn of events was made possible by an improved understanding of the biological response of living tissues to implants as well as clinical trials that validated the long-term success of these implants. Despite major structural differences between teeth and implants, such as the absence of a periodontal ligament around implants, the latter appear to provide a reliable functional replacement for their natural counterparts. This review briefly summarizes the major structural differences of the interfacial region of teeth and dental implants and their supporting tissues. It focuses on our current understanding of the soft and hard tissue responses to submerged and nonsubmerged root-form dental implants. The influence of a number of factors that affect the tissue response is reviewed, including biomaterials, implant design, surgical technique, and the local microbiota. Our recently acquired ability to modulate wound healing with guided tissue regeneration and growth factors will undoubtedly play an important role in the future utilization and success rates of dental implants. 0 1996 Wiley-Liss, Inc. Key words: Soft tissue, Hard tissue, Dental implants, Junctional epithelium, Tissue interface It is only in the last two decades that dental implants have become predictably successful for the replacement of teeth lost to trauma or disease. This success is due in a large measure to the abandonment of the trial and error approach to implant development and its replacement by a strategy based on sound biology, an understanding of materials science, and well-conceived clinical trials. As a result, data were generated that demonstrated the reliability of artificial tooth root substitutes implanted directly into bone, (i.e., endosseous implants). The geometric growth in the utilization of implants worldwide is testimony to the impact this treatment modality is having on modern dentistry. The main reason for the success of contemporary dental implants is the biocompatibility of modern implant materials with the tissues into which they are inserted. In addition, the tissues have the ability to adapt to the implants in such a way that masticatory function can be maintained, albeit with an attachment apparatus that differs structurally from that of natural teeth. When failures occur, they are generally the result of errors in clinical judgment rather than problems intrinsic to the biology of the implant-tissue interface. A brief summary of the major differences between the biological structure and function of the supporting tissues of teeth and implants will be presented. The major focus of this review will be on our current understanding of the interactions in the interfacial zone between root-form dental implants and the surrounding (or periimplant) tissues. 0 1996 WILEY-LISS. INC. SUMMARY OF DIFFERENCES BETWEEN PERIODONTAL AND PERIIMPLANT TISSUES In natural teeth the structural continuity of tooth and periodontal tissues in the recently erupted, healthy dental unit is the product of a well-integrated series of developmental events. These begin with interactions between ectodermal and endodermal components that lead to the formation of the tooth per se (i.e., dentin, enamel, and pulp). The tooth follicle that surrounds the developing tooth gives rise to the periodontal structures (i.e., the periodontal ligament, alveolar bone proper, and cementum). The developmental process leads to the intimate apposition on dentin of a cementum layer in which are embedded collagenous fibers that are continuous with the principal fibers of the periodontal ligament. The alveolar bone proper, which anchors the other end of the periodontal fibers, forms the inner lining of the dental alveolus which houses the root of the functioning tooth (for additional details, see reviews by Schroeder 1986, 1991). The supraalveolar portion of the fully erupted, functional tooth is surrounded by the soft tissues of the gingiva that are attached t o the tooth by epithelial and connective tissue elements, namely the junctional epithelium and dentogingival fibers (Fig. 1A). The outer Address reprint requests to Max A. Listgarten, University of Pennsylvania, School of Dental Medicine, 4001 Spruce Street. Philadelphia, PA 19104. TISSUE RESPONSE TO DENTAL IMPLANTS A B Fig. 1 . A Diagrammatic illustration of the interfacial region between a tooth and the periodontal tissues. The dentin surface is marked by thick arrows. The gingiva is attached to the enamel (ES, enamel space) by junctional epithelium (JE)and dento-gingivalfibers (DGF) which extend from the gingival connective tissue (CT) into the cementum layer (C) that covers the root. The primary anchorage is through the periodontal ligament (PL), the fibers of which are embedded in cementum (C) and the bone of the alveolar process (AP). AM, alveolar mucosa; GS, gingival sulcus; OE, oral epithelium. B Corresponding view of an implant, the surface of which is marked by arrowheads. The relationship of the gingiva to the implant is similar to that in A, except that dento-gingival fibers are absent. Anchorage is provided by direct contact (or osseointegration)of the implant surface with the bone of the alveolar process (AP). epithelial lining of the gingiva (oral gingival epithelium) is continuous with the sulcular epithelium that lines the shallow sulcus between the marginal gingiva and the tooth. Both of these stratified, squamous epithelial linings demonstrate varying degrees of keratinization in the absence of inflammation. They are continuous with the nonkeratinizing junctional epithelium which forms the most coronal junction of the gingiva to the tooth (Listgarten, 1972a; Schroeder and Listgarten, 1977;Schroeder, 1986). The dentogingival fibers, which are anchored in the cementum covering the cervical portion of the root, provide a firm connection between the gingiva and the tooth (Hassell, 1993). The maintenance of the structural integrity of the periodontal tissues throughout life is dependent not only on their proper development, as summarized above, but on the ability of the body to deal effectively with the wear and tear to which these structures are exposed on a daily basis. The maintenance of the attachment apparatus requires effective defenses against physical and chemical injuries and the ability to deal with infections from the ever-present oral microbiota. The intact gingival epithelial lining and junctional epithelium serve as the primary lines of defense against physical injury and bacterial invasion. Salivary secretions, with their lubricating and antimicrobial properties, increase the effectiveness of the epithelial barrier in preventing damage to the deeper structures. Should 411 the epithelium be breached, an effective inflammatory response is generally able to localize injuries to the superficial portion of the periodontium and clear the tissues of invading bacteria. The inflammatory reaction not only limits the extent of tissue damage but also assists in the mobilization of various cells needed to regenerate lost or injured tissues or at least repair structures that cannot be regenerated. For this process to work, a comprehensive inventory of cells must be available that can differentiate into the various cell types that normally inhabit this region (McCulloch, 1993). In addition, a complex communications network must be in place that allows cells to perform their various functions in an integrated and effective manner. The chemical mediators of these communications within and among cells are currently the subject of intensive investigations. They include not only traditional endocrine secretions but a host of mediators of the inflammatory process and growth factors (Kiritsy and Lynch, 1993) now generally included under the broad category of cytokines. The product of various cells, cytokines act by binding to specific receptors on cell surfaces, thereby triggering assorted biological responses that modulate the structure and function of the regional tissues (O’Neal et al., 1992; Hefti, 1993). An ever-expanding array of matrix metalloproteinases (Birkedal-Hansen et al., 1993) and their inhibitors, secreted in response to cytokine activity, serve as cellular tools that alter the structure and composition of the intercellular matrix. In contrast to natural teeth which develop with and within the tissues that will eventually provide them with their periodontal support, endosseous implants are introduced as artificial structures into a site which is surgically created within mature tissues. Consequently, the periimplant tissue support that permits the implant to function in a masticatory capacity is the result of a wound healing process rather than a developmental one. The soft and hard tissues around the implant will react to the surgical injury in a manner that will depend on a number of factors, including the nature of the cells a t the implant site, the physical and chemical nature of the implants, and the magnitude of occlusal forces applied to the implants at various stages of the healing process and later during masticatory function. In the absence of good oral hygiene, the oral microbiota may give rise to a chronic inflammation of the periimplant tissues, or periimplantitis, analogous to periodontitis around teeth. Like periodontitis, periimplantitis can lead to implant loss. Therefore, good oral hygiene is as important for the long-term retention of dental implants, as it is for the preservation of the natural dentition. The major distinguishing features between the toothl periodontium and the implant/periimplant tissue interfaces is the absence around most implants of a true periodontal ligament and the cementum layer that anchors the principal fibers of the gingiva and periodontal ligament to the implant (Fig. 1) (Listgarten et al., 1991; Donley and Gillette, 1991). The periodontal ligament contains the cellular elements that make it possible for teeth to move within their bony envelope. Cells in the ligament are responsible for the coordinated cycles of bone resorption and bone deposition 412 M.A. LISTGARTEN that result in tooth movement when physiologic forces are applied to teeth, either by the surrounding soft tissues or the orthodontist. Because implants lack this ligament, forces may be applied to them without noticeable effect on their position within the jaw. For this reason, implants can be successfully used as anchorage for orthodontic tooth movement (Roberts et al., 1989; Wehrbein and Diedrich, 1993). How the periimplant tissues react to the surgical placement of an endosseous implant t o form a more or less stable interface with this artificial object will be covered in greater detail in the following sections. Submerged (Two-Stage) Implants Because it is impossible to totally sterilize the oral cavity, some bacterial contamination of the surgical site is unavoidable. In general, microorganisms derived from the resident microbiota that contaminate a surgical site after surgical flap closure are readily GENERAL RESPONSE OF SOFT TISSUES TO DENTAL IMPLANTS This review will limit itself to a discussion of the tissue response to root-form endosseous dental implants, since they are the most commonly used. They can be divided into two major types, submerged (or two-stage) and nonsubmerged (or one-stage) implants (Fig. 2). Both types require a surgical entry and preparation of the recipient site prior to placement of the implant fixture. In the case of submerged or two-stage implants, the implant fixture is surgically inserted into bone as part of a first-stage surgical operation and completely covered with the mucosal lining postoperatively, thereby allowing the tissues to heal around the implant in a protected, bacteria-free environment (Fig. 2A). Following a suitable healing period, a secondstage surgical procedure is required to connect the submerged fixture to the oral environment by means of a transmucosal abutment (Fig. 2B). The fixture and the abutment are connected to one another, usually with a screw. Nonsubmerged or one-stage implants are inserted in a single-stage surgical procedure so that the most coronal portion of the implant immediately protrudes through the mucosal lining, making a secondstage intervention unnecessary (Fig. 2C,D). The positioning of the implant, with its occlusal surface approximately flush with the mucosal surface, prevents premature masticatory function from interfering with the healing process. However, careful hygiene to control potential postoperative infections of dental plaque origin is required (Bauman et al., 1992; Schou et al., 1992). Available data indicate that the success rates of one- and two-stage implants are comparable (Babbush and Shimura, 1993; Buser et al., 1991b; Adell et al., 1981). A ~ Fig. 2. Diagrammatic illustrations of submerged (two-stage) and nonsubmerged (one-stage) implants in a mesio-distal plane. A: Submerged implant with cover screw (CS) during healing phase. A microgap (MG) is located near the crest of the alveolar process (AP), between the cover screw and the implant fixture (IF). CT, gingival connective tissue; JE, junctional epithelium; OE, oral epithelium. B: Submerged implant after connection to the restoration (R) which, in this case, also acts as the transmucosal connector. The junctional epithelium (JE) extends to the micro-gap located near the alveolar crest. C: Nonsubmerged implant during healing phase. The cover screw (CS)is exposed and the micro-gap (MG) between the cover screw and the implant fixture (IF) is located at or near the tissue surface, well clear of the alveolar process (AP). D Nonsubmerged implant after placement of the restoration (R).The micro-gap (MG) remains close to the tissue surface. Apical migration of the junctional epithelium (JE) is minimal, with a substantial connective tissue collar (CT) between the junctional epithelium and the crest of the alveolar process (AP). L n Y TISSUE RESPONSE TO DENTAL IMPLANTS 413 cleared by the normal host defenses. However, the oc- case with the dentogingival junction, is mediated by clusal portion of the implant fixture that will be con- junctional epithelium coronally and connective tissue nected later with the transmucosal abutment may trap apically. Comparative histometric studies of noninbacteria during the insertion process (Quirynen and flamed implants and teeth, in dogs (Berglundh et al., van Steenberghe, 1993; Quirynen et al., 1994). Despite 1991, 1992) and monkeys (Lang et al., 19931, indicate the presence of a cover screw, which protects the screw few differences in the distribution of the surrounding hole from being penetrated by tissue components (Fig. soft tissue layers. The similarity in the soft tissue mea2A), bacteria under the cover screw may proliferate surements around teeth and implants is maintained and cause a chronic inflammatory reaction during the even in the presence of early gingivitis (Berglundh et initial healing phase. To minimize this potential com- al., 1992). plication, special attention must be paid to flushing out The unavoidable presence of microorganisms in the any trapped bacteria a t the time of surgery. Applica- sulcus region of teeth or implants usually results in a tion of an antibiotic ointment to the screw hole has also mild round cell inflammatory infiltrate that is localbeen recommended to minimize postoperative infec- ized to the lamina propria adjacent to the sulcus and tions (Babbush et al., 1987). Contamination through junctional epithelium. The presence of such an infilthe surgical incision is unlikely, since access is gener- trate does not necessarily lead to loss of attachment. It ally gained through some type of envelope flap, which probably reflects a normal host response to the presis designed so that the incision is located away from the ence of the resident microbiota. Should excessive bacimplant. terial deposits accumulate in the sulcus region-for exLittle published information is available on the his- ample, as a result of neglected oral hygiene-loss of tological characteristics of the healing mucosa over attachment due to chronic inflammation could occur. submerged implants. Presumably, in the absence of Such loss of attachment could also result from changes significant infection originating from beneath the in the composition of the microbiota, including an incover screw, the initial clot is resorbed, and a dense creased number of virulent species. collagenous lamina propria is formed over the implant If a gingivitis is experimentally induced in dogs and which lies more or less flush with the level of the al- allowed to progress for 90 days, the inflammatory inveolar crest. In the case of protruding implants or in- filtrate tends to extend further apically around imadequate circulation to the overlying mucosal flap, plants than around teeth with gingivitis (Ericsson et premature exposure of the occlusal portion of the im- al., 1992). The difference in the volume and apical explant is possible, as a result of localized tissue necrosis. tension of the inflammatory infiltrate is even more In such a case, a junctional epithelium presumably de- marked in ligature-induced periodontitis and periimvelops from undifferentiated basal cells of the adjacent plantitis (Lindhe et al., 1992). epithelium. We know that oral epithelium can give rise Ligature-induced periodontitis and periimplantitis to junctional epithelium from studies of wound healing have been used as models to study the destruction of following gingivectomy around natural teeth (Listgar- periodontal and periimplant tissues under controlled ten, 197213). Premature exposure of a stable, sub- conditions. The lesions result primarily from an merged implant does not seem to affect the long-term increased bacterial load, due to the plaque-retentive prognosis of the implant, which essentially becomes ligature, and possibly qualitative changes in the microbiota. In the presence of ligature-induced inflammaconverted to a nonsubmerged implant. If the implant heals without becoming exposed, it tion in the dog model, the cellular infiltrate around must be surgically uncovered to install the transmu- implants was greater, progressed more apically, and cosal abutment, which connects the tissue-integrated caused more bone destruction than around teeth implant fixture to the intraoral environment. Many (Lindhe et al., 1992). The greater extent of breakdown clinicians prefer a flap procedure to gain access to the around implants as compared to teeth was attributed implant rather than cutting a round window through primarily to the lack of dentogingival fiber insertions the mucosa directly over the implant. A flap provides in implants (compare Fig. lA,B). However, the influbetter access and flexibility, should it become neces- ence of a long junctional epithelium extending to the sary to recontour or reposition the tissues adjacent to subgingival abutment junction cannot be excluded, at the implants, and tends to preserve more of the exist- least for two-stage implants (compare Fig. 2B,D). Unlike the above results in dogs, clinical measureing masticatory mucosa (Hertel et al., 1994). The connection of the transmucosal abutment will ments around teeth and one-stage implants in monresult in a narrow gap between the fixture and the keys were comparable before as well as after ligatureabutment, the size of which will vary according to the induced inflammation (Lang et al., 1993).The different precision with which the parts were designed and man- results between the dog and monkey models could be ufactured. This gap is usually located near the bone due to species variation. However, this is not likely, surface (Fig. 2B). Theoretically, the gap could serve as since such differences have not been reported in past a potential source of infection by bacteria that may studies of ligature-induced periodontal disease in these become trapped in it during the second-stage surgical animals. A more likely reason is the presence of a subprocedure or during the immediate postoperative pe- gingival gap in the two-stage implants used in the dog riod. However, as will be seen later, there is no evi- studies, whereas one-stage implants, consisting of a dence that bacterial contamination of the gap region in single element, were used in the monkeys. In the dogs, two-stage implants tended to show a two-stage implants has any effect on probing measurements, as compared to one-stage implants that have no greater thickness of supracrestal connective tissue than teeth (1.7 mm vs. 1.1 mm) and parallel orientagap. The junction of the gingiva with an implant, as is the tion of the gingival connective tissue fibers to the im- 414 M.A. LISTGARTEN plant surface instead of the more or less perpendicular insertion of dentogingival fibers typical of teeth (Ericsson and Lindhe, 1993). Probing depths tended to be greater around implants than teeth (2.0 mm vs. 0.7 mm), an observation that the authors attribute to the architecture of the supracrestal connective tissue around the implants. They consider this tissue to be less resistant to the passage of a probe than the inserted dentogingival fibers around teeth. Since the implants were separated from the soft tissue specimens for histological processing, the actual tissue-implant interface was no longer intact. Therefore, it is questionable how reliably the authors could locate the position of the probe tip or the soft tissue landmarks in relation to the implant surface. It is likely, for example, that the apical extent of the junctional epithelium was underestimated, due to epithelial tears a t the tissue-implant interface, with portions of the junctional epithelium remaining attached to the discarded fixture. Schou et al. (19931, who worked with a monkey model, came to the conclusion that the lack of a periodontal ligament could also be a factor in the relative susceptibility of periimplant tissues to ligature-induced breakdown. Their conclusion, however, is open to challenge. First, they experimentally created ankylosed teeth that lack a periodontal ligament by extracting and reimplanting teeth after drying the roots. They then induced chronic inflammation with ligatures around normal teeth, ankylosed teeth, and titaniumcoated polycarbonate implants. After 7 weeks, significant bone loss was observed around implants and ankylosed teeth but not around nonankylosed teeth. However, the bone loss was considerably greater around the implants than around the ankylosed teeth. No convincing explanation was given to explain these findings. It is possible that reimplantation of the teeth to achieve ankylosis did not result in the restoration of a dentogingival connective tissue junction with the same quality as the original, particularly with respect to the density of dentogingival fiber insertions. Therefore, while the reimplanted teeth may have been more resistant to breakdown than implants, they failed to match the resistance of the undisturbed teeth. Thus, the differences among the three experimental groups is more likely attributable to differences in the structure of the dentogingival junction than the mere presence or absence of a periodontal ligament. Adell et al. (1986) reported that in relatively inflammation-free implants placed in humans, some decrease in probing depth takes place after 6 months postinsertion, from an initial depth of 3.8 mm to an average probing depths ranging from 2.4-2.9 mm. The original probing depth likely represents mean measurements taken close to the alveolar crest, whereas the shallower measurements reflect decreased probe penetration, probably due to increased connective tissue density and decreased inflammation. Lekholm et al. (1986) reported an average probing depth of 3.8 mm for 125 fixtures placed in 20 patients, with the greater measurements associated with inflammatory changes. The depth of the measurements suggests that they were taken with the probe tip close to the alveolar crest. Unlike the situation in most animal experiments, two-stage implant abutments in humans are changed a t least once, namely when the healing abutments are replaced by the abutment incorporated in the restoration (Fig. 2B). Changing abutments probably favors the development of a long junctional epithelium extending close to the gap between fixture and abutment, a situation which differs from that in the dog model, where the original abutment is left in place (Ericsson and Lindhe, 1993; Berglundh et al., 1991). A long junctional epithelium in conjunction with an adjacent inflammatory infiltrate would favor more apical probing measurements. Apse et al. (1989) measured probing depths around teeth and two-stage implants in partially edentulous patients with similar plaque and gingival indices. Implants routinely gave mean probing depths greater than those of natural teeth by a magnitude of around 0.8 mm (3.29 vs. 2.47). The discrepancy between the two is likely due to a longer junctional epithelium around the implants, extending close to the gap, and possibly a more apical extension of the inflammatory infiltrate not detectable on visual inspection. When wound healing is completed, following second stage surgery, the mucosa should be tightly adapted to the abutment. Many clinicians prefer the periimplant gingiva to consist of fixed masticatory rather than moveable lining mucosa (Adell et al., 1981; Buser, 1987; ten Bruggenkate et al., 1990; Artzi et al., 1993). However, Wennstrom et al. (1994) reported no correlation between the periodontal status of implants, in function for a t least 5 years, and the width of the surrounding masticatory mucosa. Nonsubmerged (One-Stage) lmpiants The healing of the soft tissues following the insertion of one-stage implants resembles that which takes place around transmucosal abutments after second stage surgery (Figs. 2,3). The major differences between the two types of implants is the absence in one-stage implants of the gap between the implant fixture and the transmucosal abutment, which is typical of two-stage systems, and the transmucosal location of the onestage implant immediately after insertion. While the trapping of bacterial contaminants in the supracrestal region is much less likely with nonsubmerged systems that lack the subgingival gap, much greater postoperative care is required t o prevent gingival infection during initial healing, since these implants are immediately exposed to the oral environment (Gotfredsen et al., 1991). Following insertion of the fixture and readaptation of the soft tissues to the implant, a clot fills the residual voids between the soft tissues and the fixture. Organization of the clot is accomplished by ingrowth of granulation tissue from the surgical flap and the adjacent bone. In the case of periodontal surgery, granulation tissue also originates from the periodontal ligament. This source of granulation tissue is obviously missing around conventional implants where the recipient site is drilled directly into edentulous bone. It is the absence of granulation tissue of periodontal ligament origin that accounts for the major structural difference in the anchorage between teeth and implants to the adjacent bone. Under experimental conditions, it is possible to generate a tooth-like attachment apparatus around im- TISSUE RESPONSE TO DENTAL IMPLANTS Fig. 3.A Histologic section through nonsubmerged (one-stage)implant replica made from titanium-coated epoxy resin, 3 months after insertion in a dog mandible. B = bone of alveolar process; G = gingiva; I = implant. Bar = 1mm. B: Magnified view of alveolar crest (AC) region of A. Note minimal sulcus formation and minimal apical migration of junctional epithelium (single arrow). With the exception of a shallow groove in the crestal region, the alveolar process is in direct contact with the titanium layer coating the implant replica (double arrow). Bar = 0.2 mm. C Magnified view of the gingival 415 region in A. The arrow indicates the junction of the smooth collar region (above arrow) with the rough-surfaced apical region. The connective tissue fibers run predominantly in a plane parallel to the titanium-coated implant surface. Bar = 50 pm. D Magnified view of A. Bone (B) surrounding the implant replica is in direct contact with the titanium coating (arrows). Bar = 0.2 mm. (Figs 3A and 3B Reproduced from Listgarten et al., 1992, with permission of the publisher.) 416 M.A. LISTGARTEN plants, with a cementum layer adherent to the implant surface, anchoring the connective tissue fibers of a periodontal-like ligament. To achieve this unusual healing pattern, implants were purposefully inserted into recipient sites adjacent to residual root tips still surrounded by their periodontal ligament. Progenitor cells in the granulation tissue derived from the periodontal ligament migrated to the implant surface and were able to reconstitute a tooth-like attachment apparatus around the implant, with a cementum layer adherent to the implant surface and ligament fibers embedded in it (Buser et al., 1990a,b; Warrer et al., 1993). These highly specialized cells are not found in the granulation tissue that participates in the normal healing process after implant placement. Therefore, this healing pattern represents a very atypical one for implants. Nevertheless, the observation is of great biologic interest, as it highlights the unique cellular contribution of the periodontal ligament t o wound healing. On the basis of periodontal wound healing studies (Listgarten et al., 19821, we know that following the repositioning of surgical flaps, epithelium begins to proliferate from the cut gingival margin along the inner surface of the flap. It migrates as a thin sheet of cells between the newly organized clot lining the flap and nonorganized remnants of the clot and other debris located on the root surface. When the epithelial cells reach the granulation tissue that binds the basal portion of the flap to the tooth, they stop their apical migration, become attached to the tooth, and begin to migrate coronally. In the process they clear the wound of accumulated debris and reestablish a n organic connection between the gingiva and the root surface by means of a long junctional epithelium. Presumably, a similar process occurs following the surgical placement of one-stage implants. Precisely drilled receptor sites will insure a tight fit at the crest of the preparation, thereby allowing the granulation tissue to rapidly bridge the small space left between the bone and the implant. The proliferating epithelium will thus be prevented from growing in between the implant and the bone. It is to be expected that in the first 3 weeks after one-stage implant placement, or in the case of two-stage implants after abutment connection, a long junctional epithelium will become established, which will extend from the base of the gingival sulcus to a level close to the alveolar crest. It bears emphasizing that the formation of the junctional epithelium proceeds in an apico-coronal direction, from epithelium that initially migrated from the gingival margin toward the base of the flap. In vitro data indicate that epithelial cells migrate more readily on smooth- than rough-surfaced titanium (Cochran et al., 1994). The epithelial cells adjacent to the implant surface form typical hemidesmosomes and a basal lamina which mediate their attachment to the surface (James and Schultz, 1974; Listgarten and Lai, 1975; Bauman et al., 1993) yet allow the cells to move coronally. The extent to which supracrestal connective tissue eventually replaces the apical portion of the junctional epithelium (Listgarten et al., 1982) will depend in part on the degree to which inflammation in the supracrestal region can be controlled. The granulation tissue in the supracrestal region eventually becomes remodeled into a dense collagenous connective tissue with the majority of fibers running in a direction more or less parallel to the implant surface (Figs. lB, 2D, 3C) (Buser et al., 1992; Listgarten et al., 1991, 1992). In a histologic study in dogs, Buser et al. (1992) reported limited apical extension of the junctional epithelium. Immediately apical to the epithelium, next to the implant, they reported a 50-100 p,m wide supracrestal connective tissue collar composed of dense, scar-like, relatively avascular connective tissue. The presence of connective tissue fibers running parallel to the implant surface is the usual finding around most types of implants. A few reports indicate the possibility of perpendicular fiber insertions in the case of rough-surfaced, porous implants (Schroeder et al., 1981; Deporter et al., 1988; Buser et al., 1989; Pilliar, 1991). It is evident, however, that perpendicular fiber insertions are rare and are not a common finding following the placement of most implant types, even when they have relatively rough but nonporous surfaces (Buser et al., 1991a; Listgarten et al., 1992). Because of the gap between abutment and fixture in two-stage implants, inflammation following abutment connection is theoretically more likely to persist around two-stage than one-stage implants. The absence of this potential bacterial trap around one-stage implants should facilitate the undisturbed organization of the postsurgical clot and the formation of a wider collar of supracrestal connective tissue than would be the case with two-stage implants. Therefore, one might expect deeper probing depths to persist around two-stage than one-stage implants. Yet this is by no means obvious from published clinical measurements. Buser et al. (199Oc, 1991b) reported average probing depths around one-stage implants of 2.69-2.81 mm between 1 and 2 years after implantation. This is shallower than the 3.8 mm reported by Lekholm et al. (1986) for two-stage implants but is similar to average probing depths of 2.89 mm reported by Apse et al. (1989) and the mean measurements of 2.4-2.9 mm observed by Adell et al. (1986) after tissue healing was complete. Therefore, there is no detectable difference between one- and two-stage implants with respect to clinical probing depth measurements after they have been in function for several months or longer. Definitive histological evidence, based on block sections of intact tissue-implant interfaces, is still lacking to show what differences, if any, exist in the comparative healing of periimplant tissues to one- and two-stage implants. Addendum Although implants were separated from their surrounding tissues prior to processing, Abrahamsson et al., (1996)have recently demonstrated similar tissue to implant relationships in 1-stage and 2-stage implants, 6 months following abutment installation. Control of Epithelial Migration Junctional epithelial proliferation is a characteristic feature of periodontal sites that are inflamed (Listgarten, 1986; Page and Schroeder, 1982). In health, the junctional epithelium has a relatively smooth external basal lamina and occupies a relatively stable position on the tooth surface. In the presence of inflammation, it develops numerous papillary extensions which in- TISSUE RESPONSE TO DENTAL IMPLANTS vade the adjacent inflamed connective tissue. The junctional epithelium also tends to proliferate in an apical direction as dentogingival fibers are destroyed. The absence of dentogingival fibers around implants has been blamed for the more rapid apical extension of the inflammatory process around implants as compared to teeth (Lindhe et al., 1992), a process which may also promote apical proliferation of the junctional epithelium. Inflammation may also affect the expression of various cytokeratins in the gingival epithelia (Bosch et al., 1989) and generate a variety of cytokines that can influence the behavior of adjacent epithelial cells (MacKenzie, 1988). Mackenzie et al. (1991) have reported that different portions of the gingival connective tissue may differentially affect the expression of cytokeratins in human gingival epithelia. In an earlier review of mesenchymavepithelial interactions, MacKenzie and Hill (1984) stated that connective tissues play a key role in controlling epithelial morphogenesis and cytodifferentiation. Given the importance of connective tissue controls on epithelium, it is likely that gingival connective tissues play an important role in determining the degree of apical proliferation of the junctional epithelium. One factor which may explain the limited downgrowth of mature junctional epithelium around teeth and implants may be related to differences in the interactions between epithelial cells and various constituents of the gingival connective tissue, namely the lamina propria and what has been described by Mackenzie and Hill (1984) and Mackenzie and Fusenig (1983) as “deep” connective tissues. Deep connective tissue beds appear to lack the physical structure andlor the diffusible substances required to stimulate epithelial cell migration and differentiation. Whereas the lamina propria of the gingiva will stimulate the overlying epithelium to proliferate and maintain a multilayered, keratinizing phenotype, deep connective tissues may inhibit both the proliferation as well as the differentiation of gingival epithelium. If the deeper portions of the gingival connective tissues share the same properties as the deep connective tissues in Mackenzie’sexperiments, epithelial migration could be inhibited when the advancing edge reaches this region. Thus, in addition to other factors discussed previously, the apical extent of the junctional epithelium may be determined by the presence in gingiva of deep connective tissues that inhibit epithelial migration. Presumably, similar controlling mechanisms are operational around both implants and teeth. 417 the long term, osseointegration is considered to be more predictably successful than a fibrous union. Brhnemark et al. (1977)described osseointegration as a relationship where “bone tissue is in direct contact with the implant, without any intermediate connective tissue.” Later, Brinemark et al. (1985) stated that osseointegration represents “a direct structural and functional connection between ordered, living bone and the surface of a load-carrying implant.” This latter definition fails to include healing of bone to implants that are not in function yet have all the morphologic characteristics of being osseointegrated by the first definitionfor example, implants in situ-prior to abutment connection. The definition has been repeatedly reinterpreted. Perhaps one of the more extreme reinterpretations states that since “there is no acceptable histologic correlate to the term ‘osseointegration,’ ” implants that healed with a histologically distinct fibrous union but were clinically stable also should be considered as osseointegrated (Nystrom et al. 1993). In order to minimize further confusion, the term osseointegration, as used in this review, describes a direct contact between living bone and the implant surface, based on histological evidence of the type available around the time the definition was coined (Schroeder et al., 1976; Brinemark et al. 1977). While clinical signs, particularly with dynamic percussion devices (Teerlink et al., 19911, may suggest that osseointegration has taken place, the interpretation of these signs as evidence of osseointegration can only be presumptive, since, with current technology, only histological sections can provide definite proof of the absence of fibrous tissue between the bone and the implant. The nature of the implant to bone interface depends on numerous factors, including the type of implant (Smith, 19931,the surgical technique and quality of the surrounding bone structure, the timing and magnitude of implant loading, and the degree of postsurgical oral hygiene. Type of Implant Today most root-form implants are composed of materials that will allow healing by osseointegration, with titanium being the most common. Others generally known to heal by osseointegration include hydroxyapatite and aluminium oxide ceramics (Stef lik et al., 1992a,b) and zirconium (Albrektsson et al., 1985; Akagawa et al., 1993).It has been amply demonstrated that a strong bond can become established between bone and implants made of commercially pure grade titanium (Brinemark et al., 1977; Schroeder et al., GENERAL RESPONSE OF OSSEOUS TISSUE TO 1981, 1991; Albrektsson et al., 1983). However, titaDENTAL IMPLANTS nium does not have the ability to induce osteogenesis The healing of bone tissue around endosseous im- from potential osteogenic precursor cells in marrow plants has been the subject of a large number of pub- (Rahal et al., 19931,nor does it possess osteoconductive lications (for recent reviews see Listgarten et al., 1991; properties (i.e., the ability to induce bone to grow along Bidez and Misch, 1992; Roberts et al., 1992; Kohn, its surface). Given the heterogeneity in bone structure 1992; Albrektsson and Zarb, 1993; De Lange and De at the various sites receiving titanium implants, a Putter, 1993; Steflik et al., 1993a). Unlike teeth, which great deal of variation in bone contact can be expected are anchored to the alveolar bone by a periodontal lig- (Ettinger et al., 1993). By contrast, hydroxyapatite ceramic implants and ament, implants placed in bony sites heal either by forming a direct and intimate contact to bone, com- hydroxyapatite coatings exhibit osteoconductive propmonly referred t o as osseointegration (Brinemark et erties (De Lange and de Putter, 1993; Weinlaender et al., 1977,1985), or a looser union mediated by a fibrous al., 1992; Gottlander et al., 1992; Teranobu et al., capsule, or “fibrosteal integration” (Weiss, 1986). For 1989). As a result, new bone tends to spread along the 418 M.A. LISTGARTEN implant surface, thereby forming a thin layer of compact bone, even in areas where the implant is adjacent to marrow spaces. While hydroxyapatite coatings may contribute to more rapid osseointegration in the early stages of healing, they do not seem to confer any advantages over the long term. The slow resorption of hydroxyapatite coatings has been interpreted by some as placing such implants a t greater risk than plain titanium implants. However, there are no convincing data to support this belief (Zablotsky, 1992). Assorted stainless steel and chrome cobalt alloys that constitute the mainstay of orthopedic implants (Albrektsson and Hansson, 1986; Linder, 1989; Linder et al., 1989) and certain titanium alloys (Johansson et al., 1989) were thought to be incapable of healing by osseointegration. However, more recent investigations have identified some of these and other materials as having a similar ability to bind to bone, including vitallium (Niki et al., 1991; Johansson et al., 1991; Albrektsson and Johansson, 1991), niobium and various steel alloys (Albrektsson and Johansson, 1991; Johansson and Albrektsson, 1991), hydroxyapatite ceramic (Gottlander and Albrektsson, 1991; Lemons and Bidez, 1991; Smith et al., 1992; De Lange and De Putter, 1993), and bioglass (Lemons and Bidez, 1991). However, Barth et al., (1990) were unable to confirm the osseointegration of a glass ceramic they tested in parallel with titanium implants. To favor retention and distribute functional forces more evenly (Skalak, 1983; Brunski, 1991), some implant designs incorporate a screw thread in addition to the basic cylindrical form that characterizes most of the current implant models. However, the long-term success rate appears to be similar for screw-shaped and non-screw-shaped root-form implants (Buser et al., 1991b). Some designs increase the surface area available for tissue integration by using rough surfaces such as plasma-sprayed surfaces or by roughening the surface by acid-etching or sandblasting techniques. While short-term differences have been reported between percentage of bone contact (Buser et al., 1991a) and shear strength (Wilke et al., 1990; Carlsson et al., 1988) for various surface textures, there is no clear evidence of long-term superiority of one surface texture over another. It should be stated in passing that the poor track record of blade implants is not so much the result of their unique shape as of some other critical factors that were ignored in the early days of their use. These include their fabrication with materials that were not readily able to achieve osseointegration, inadequate surgical technique, and premature loading, to name a few of the main causes of failure. While long-term data are not available at this time, there is some indication that blade implants fabricated with modern materials, inserted with good surgical technique and without overheating bone, and not prematurely loaded can become osseointegrated and provide satisfactory prosthetic support (Lum et al., 1991). Surgical Technique and Bone Structure Most systems provide instrumentation that allows the surgeon to create an intrabony recipient site for the implant that will provide a tight fit for the implant fixture, thereby preventing fibrous tissue ingrowth and interference with osseointegration. The instrumentation is designed to create a suitable site with minimal trauma to the bone, thereby maintaining the vitality of most of the remaining bone cells. Preparation of the recipient site in an edentulous ridge requires special care with respect t o asepsis, avoidance of excessive trauma to the soft and hard tissues, achieving primary stability of the implant in situ, and good postoperative care. In the case of one-stage implants, additional care must be given to avoid plaque accumulation during the critical wound healing phase that immediately follows the placement of the implants. Since one-stage implants are placed in direct communication with the oral cavity, good coaptation of the flap edges and the regular use of disinfecting mouth rinses promote rapid healing of the soft tissues around the implant, thereby forming a seal that prevents bacterial ingress. Assuming that all the critical steps in the placement of an implant are followed, the wound healing process should lead to osseointegration. Continuous remodelling will preserve the intimate contact with the bone and the reparative potential needed to offset functional wear and tear (Listgarten et al., 1991; Roberts, 1993). Disinfection of the oral cavity and the use of sterile operating conditions have been considered essential for many years (Branemark et al., 1985). Recently, a comparative study of a strict operating room protocol compared to a more relaxed aseptic technique failed to reveal a statistically significant difference in the success rate of implants placed under either protocol (Sharf and Tarnow, 1993). However, the results should be interpreted with caution, since the failure rate is low with either technique and the clinical trial had relatively little statistical power to discriminate between two relatively successful techniques. Avoidance of tissue trauma is inherent to any surgical intervention, since it minimizes the risk of infection and optimizes the repair process. Bone is uniquely sensitive to environmental trauma. For example, it has been reported that heating bone to 47°C for 1 min is capable of killing osteogenic cells, thereby compromising the chances of achieving ideal healing. Therefore, particular emphasis must be placed on avoiding excessive heating of the bone by using serial drills, relatively low drilling speeds, abundant irrigation, light hand pressure, and sharp instruments (Ericsson and Albrektsson, 1983). Whether one- or two-stage implants are used, much attention must be given to achieving good primary stability at the time of implant placement and avoiding premature loading of the implant (Brunski, 1991). Both of these factors could result in sufficient micromovement to interfere with healing by osseointegration. The quality and quantity of bone at the recipient site are also key elements in achieving primary stability (Brhemark et al., 1985). There should be enough bone volume to totally surround the intrabony portion of the implant and a sufficient amount of compact bone to supply the rigid immobilization required for osseointegration. Following the insertion of the implant into the precisely cut recipient site, the fibrinous clot a t the interface should occupy a minimum volume, a situation that TISSUE RESPONSE TO DENTAL IMPLANTS will facilitate the organization and replacement of the clot by bone. A precise fit in the region of the crest is critical. Because of the close proximity of the gingival tissues to the recipient site, any space between the crestal bone and the implant is likely to become invaded by gingival rather than bone cells, with the result that fibrous tissue rather than bone will fill that space. A precise fit will also insure primary stability, a prerequisite for osseointegration. Despite good surgical technique, it is likely that a thin seam of bone adjacent to the prepared site will become nonvital along the entire recipient site. This is not critical as long as the damage is limited. First, the rate of bone resorption is relatively slow and does not progress a t an even rate over all the damaged surfaces. Therefore, it is unlikely that remodelling of the damaged bone will interfere with primary stability of the implant and osseointegration. Evidence of active remodelling close to the implant surface has been provided by bone markers administered a t specific time intervals following implant placement (Roberts et al., 1984, 1986). Considerable bone formation near the implant can be detected as early as 3 days after implant placement. However, several weeks are needed for replacement of the intervening tissue with bone at the implant surface (Roberts et al., 1984). Early bone formation is characterized by a callus-like woven bone which originates from the cut bone surface (Clokie and Warshawsky, 1995). Later in the repair process, between 6 and 16 weeks, the woven bone is remodelled into lamellar bone (Roberts et al., 1984; Clokie and Warshawsky, 1995), either as trabecular bone or osteon-containing (Haversian) bone, depending on the local bone structure (Sennerby et al., 1992). Autoradiographic and ultrastructural studies of early wound healing of experimental titanium implants in rat tibias have indicated various patterns of bone growth in relation to the implants. In addition to confirming the conversion of the trabecular bone, laid down in the early stages of healing, into more compact bone, the results indicate that bone can grow toward the implant from more distant sites as well as away from it after having started to form on the implant surface (Nanci et al., 1994). After the implant is in function, further remodelling will depend on the magnitude an direction of the applied forces. It has been shown that after implants are placed in function, remodelling results in increased bone volume and density in the cervical region, as compared to unloaded control implants (Piatelli et al., 1993a). Interestingly, in this particular study the percentage bone contact did not differ between loaded and unloaded implants. It is clear that cancellous bone with large marrow spaces does not provide the same physical support for a freshly inserted implant as that provided by compact bone, Therefore, it is desirable to prepare the recipient site so as to include at least part of it within compact bone, thereby insuring that primary stability will be achieved and that functional forces will be more readily withstood. In general, remodeling after implant insertion will follow the pattern of the preexisting bone (Parr et al., 1993). Where compact bone was present, it will be replaced with compact bone. Where sparse tra- 419 becula contacted the bone, sparse bony contacts will be reestablished (Listgarten et al., 1992). An implant located in very cancellous bone will generally have only a small portion of its surface in contact with a few bone spicules, the rest of the surface being in contact with the nonmineralized contents of the marrow spaces. By contrast, an implant in compact bone may approach 100% bone contact, since cellular and marrow elements are sparse. However, the stability of an implant is not solely dependent on the percentage of bone contact. Greater stability can be achieved when an implant is located in compact bone as compared to one with a comparable bone contact ratio but located in cancellous bone (Sennerby et al., 1992). Interestingly, the failure rate of implants does not seem to be affected by osteoporosis (Dao et al., 1993). Histological studies of successfully osseointegrated implants have generally shown a direct contact of the surrounding bone with the implant. However, studies at the ultrastructural level have revealed much more variation a t the implant to bone interface than is evident from light microscopic studies. These reports indicate that the degree of proximity of bone to the implant surface falls within a range that extends from intimate contact between the bone and the implant (Fig. 4) Brunette et al., 1991; Listgarten et al., 1992)to the presence of nonmineralized and mineralized intervening layers of varying width and electron density (Linder et al., 1983; Albrektsson et al., 1983,1985; Albrektsson and Hansson, 1986; Ericson et al., 1991; Sennerby et al., 1991; Steflik et al., 1992a,b, 1993a,b; De Lange and De Putter, 1993; Piatelli et al., 1993b;Nanci et al., 1994). It is not clear to what extent these differences are due to true structural variations or to different methodologies employed by the investigators to examine the interfacial region. Should excessive devitalization of bone occur as a result of thermal injury, osseointegration will be compromised, and a fibrous capsule is likely to form between the implant surface and the bone. A similar reaction is likely if the implant does not achieve primary stability or is loaded prematurely, before the bone has become sufficiently remodelled around it to prevent micromotion during functional movements (Brunski, 1991).Fibrous capsules may also form around implants made of materials that do not readily osseointegrate, such as vitreous carbon. The presence of fibrous capsules after healing is not considered acceptable for dental implants, since implants that heal in this manner appear t o have a less predictable and less successful track record than osseointegrated implants (Albrektsson and Sennerby, 1990). Such fibrous layers should not be confused with the thin, acellular layers reported in the interfacial region of bone and implants. These layers, resembling a lamina limitans, are composed of glycoconjugates such as osteopontin (Nanci et al., 1994) and do not affect the long-term survival of implants, as fibrous capsules might. They are fully compatible with the structural descriptions of osseointegrated implants and their reported long-term success rates. These layers appear similar in structure and composition to the cement lines seen at bone interfaces as a result of cyclic deposition and remodelling (McKee and Nanci, 1993). Placing implants directly into bone grafts remains a 420 M.A. LISTGARTEN Fig. 4. Transmission electron micrograph of the interface between osseointegrated, titanium-coated epoxy resin replica (T,titanium coating) and the adjacent bone (B). The section was made after demineralization of the tissue block which reveals the individual collagen fibers (arrowheads)of the bone matrix. Note the intimate contact between the titanium surface and the individual collagen fibers of the bone matrix. Bar = 0.25 pm. questionable procedure, even with an ideal grafting material such as the iliac crest autograft, since the implants appear to heal by fibrous encapsulation (Nystrom et al., 1993). This result may be due to the relatively great distance which exists between the implant and the closest source of well-vascularized bone. A similar problem may exist with implants placed in recipient sites generated by sinus elevation procedures. Histologic studies of bone cores obtained from such sites indicate retention of allografted elements with very sparse ingrowth of host bone, even 6 months postoperatively, although the results with autogenous bone grafts seem more promising (Nishibori et al., 1994). GUIDED TISSUE REGENERATION (GTR) In order to promote the ingrowth of desirable progenitor cells into a wound and keep out undesirable cell types, a variety of physical cell barriers have been devised. These devices, usually in the form of membranes, help to create or maintain tissue compartments and promote granulation of the space with desirable cells (O’Neal et al. 1994). Originally devised to regen- erate lost attachment around periodontally diseased teeth (Karring et al., 19931, guided tissue regeneration (GTR) has found numerous applications, including the augmentation of alveolar ridge dimensions prior to or in conjunction with implant placement (Seibert and Nyman, 1990; Dahlin et al., 1991; Lang et al., 1994b; Schenk et al., 1994) and the placement of implants into recent extraction sockets (Warrer et al., 1991; Gotfredsen et al., 1993; Lekholm et al., 1993; Becker et al., 1994; Gher et al., 1994a,b; Lang et al., 1994a). GTR has also been reported to be successful in achieving osseointegration to surgically as well as pathologically denuded implant surfaces and to titanium as well as hydroxyapatite-coated fixtures (Zablotsky et al., 1991; Jovanovic et al., 1993; Caudill, 1993; Lundgren et al., 1994). To facilitate the maintenance of a compartment for tissue ingrowth, titanium-reinforced membranes have been developed which appear to be particularly helpful in ridge augmentation procedures (Jovanovic et al., 1995). However, therapeutic outcomes with GTR are not consistent or predictable, and premature exposure of the barrier, a relatively common occurrence, as well as infection around the membranes can severely compromise the degree of tissue regeneration (Warrer et al., 1991; Gotfredsen et al., 1993; Becker et al., 1994; Lang et al., 1994b; Rominger and Triplett, 1994). Schenk et al. (1994) have provided some excellent histological data on the regeneration of large ridge defects. The importance of maintaining the integrity of the tissue compartment in which regeneration is expected is well demonstrated. The results also clearly show how the new bone is derived from the osseous surfaces adjacent to the defect and undergoes a maturation process leading to a denser trabecular network. The structural reorganization of the lamellar bone is reminiscent of that of normal membrane bone development and maturation. Similar observations were reported by Hammerle et al. (1995) in healing calvarial defects produced in a rabbit model. In one animal experiment, in which different types of dental implants were intentionally stripped of bone and contaminated with dental plaque, results have shown that regeneration of lost bone is possible. Osseointegration could be demonstrated in discrete portions of the surface following cleansing of the implant surfaces with a powder abrasive and citric acid, followed by guided tissue regeneration. The authors reported greater bone contact to hydroxyapatite-coated than titanium implants (Jovanovic et al., 1993). On the other hand, healing with fibrous tissue was the predominant healing pattern in a similar study, but without the use of citric acid. In another study, little difference in the healing pattern was detected between implants treated with guided tissue regeneration and those treated by conventional debridement (Schupbach et al., 1994). Even in noncontaminated dehiscences, intentionally produced at the time of implant insertion, guided tissue regeneration resulted in covering the defects with a bone-like tissue which, however, was not in intimate contact with the implant surface (Palmer et al., 1994). While clinical studies in human subjects suggest a beneficial effect of GTR on the repair of periimplant defects, the studies often lack adequate controls (Jovanovic et al., 1992; Shanaman, 1994) and reproducibility by other investigators. 421 TISSUE RESPONSE TO DENTAL IMPLANTS crepant findings may be due in part to differences in the growth factors tested, their dosages and mode of In order to optimize the ingrowth of desirable cells administration, and differences in the experimental deinto tissue compartments created by GTR, attempts signs and animal models. The application of growth factors to improve wound have been made to combine the procedure with topical applications of growth factors (Lynch et al., 1991; healing around implants is still in an embryonic stage, Becker et al., 1992; Amar and Chung, 1994; Graves and much more experimental work is needed to deterand Cochran, 1994;Wang et al., 1994).Platelet-derived mine dosage, mode of application, sequence of delivery, growth factor (PDGF) and insulin-like growth factor-I and optimal combinations that will achieve the desired (IGF-I) are found in bone and have the ability not only aims. Until now, some of this work has been hampered to stimulate osteoblastic activity but also to act as by the cost and lack of availability of these agents. chemoattractant in order to recruit important cells to However, modern production techniques should allevithe wounded area (Kiritsy and Lynch, 1993; Lynch, ate both the cost and shortage of these biologically pow1994). When applied topically to fenestration defects in erful agents. Eventually, the benefit of growth factors dogs, PDGF stimulated periodontal ligament fibroblast on wound healing around implants will have to be asproliferation, regardless of the presence of a barrier sessed in well-controlled clinical trials. membrane (Wang et al., 1994). In combination, PDGF LITERATURE CITED and IGF-I appear to promote the formation of a denser trabecular network around recently inserted implants Abrahamsson, I., T. Berglundh, J. Wennstrom, and J. Lindhe. 1996 The periimplant hard and soft tissues at different implant systhan comparable controls, with a greater percentage of tems. A comparative study in the dog. Clin. Oral Implants Res., bone contacts on the implants (Lynch et al., 1991; in press. Becker et al., 1992). PDGF, a dimeric protein consist- Adell, R., U. Lekholm, B. Rockler, and P.-I. Brlnemark 1981 A 15ing of two distinct polypeptides, exists in three isoyear study of osseointegrated implants in the treatment of the edentulous jaw. Int. J . Oral Surg., 10:387-416. forms, two of which are also capable of inducing prostaglandin-mediated bone resorption (Cochran et al., Adell, R., U. Lekholm, B. Rockler, P.-I. Brlnemark, J . Lindhe, B. Eriksson, and L. Sbordone 1986 Marginal tissue reactions at os1993). seointegrated titanium fixtures. I. A 3-year longitudinal prospecIn a series of reviews on regeneration (Linde et al., tive study. Int. J. Oral Maxillofac. Surg., 15:39-52. 1993; Amar and Chung, 1994; Graves and Cochran, Akagawa, Y., Y. Ichikawa, H. Nikai, and H. Tsuru 1993 Interface histology of unloaded and early loaded partially stabilized zirco1994; O”ea1 et al., 19941, the authors suggested that nia endosseous implant in initial bone healing. J . F’rosthet. Dent., the osteopromoting activity of GTR could probably be 69:599-604. enhanced with a variety of growth factors, including Albrektsson,T., P.-I. Brlnemark, H.-A. Hansson, B. Kasemo, K. Larsson, I. Lunstrom, D.H. McQueen, and R. Skalak 1983 The interbone morphogenetic proteins (BMP). Sailer and Kolb face zone of inorganic implants in vivo: Titanium implants in (1994) reported three cases in which BMP combined bone. Ann. Biomed. Eng., 11:l-27. with iliac bone grafts resulted in excellent healing Albrektsson, T., and H.-A. Hansson 1986 An ultrastructural characaround compromised implants. However, the absence terization of the interface between bone and sputtered titanium or stainless steel surfaces. Biomaterials, 7:201-205. of suitable controls and the low number of treated cases precludes any reliable generalization. By incorporating Albrektsson, T., H.-A. Hansson, and B. Ivarsson 1985 Interface analysis of titanium and zirconium bone implants. Biomaterials, bovine BMP fractions in an insoluble collagenous bone 6:97-101. matrix, that served both as a vehicle and as a spacer for Albrektsson, T., and C. Johansson 1991 Quantified bone tissue reactions to various metallic materials with reference to the so-called GTR, Ripamonti et al. (1994)reported improved regenosseointegration concept. In: The Bone-Biomaterial Interface. eration of furcation defects in baboon molars over deJ.E. Davies, ed. University of Toronto Press, Toronto, pp. 357fects treated with the matrix alone. They concluded 363. that recombinant human BMPs are potentially valu- Albrektsson, T., and L. Sennerby 1990 Direkte Verankerung von Oralen Implantaten: Klinische und experimentelle Betrachtunable adjuncts to the regeneration of periodontal tissues. gen des Konzepts der Osseointegration. Parodontologie, 1:307By means of recombinant human transforming 320. growth factor (31 (rhTGF-Pl) applied locally to skull Albrektsson, T., and G.A. Zarb 1993 Current interpretations of the defects in rabbits, Beck et al. (1993) were able to demosseointegrated response: Clinical significance. Int. J. Prosthodont., 6:95-105. onstrate the ability of rhTGF-(31 to stimulate the reS., and K.M. Chung 1994 Clinical implications of cellular bicruitment and proliferation of local osteoblasts, al- Amar, ologic advances in periodontal regeneration. Curr. Opin. Perithough remodelling rates were not altered. Similar odontol., 128-140. findings were obtained by Aufdemorte et al. (1993) by Awe. P., R.P. Ellen. C.M. Overall. and G.A. Zarb 1989 Microbiota and crevicular fluid collagenase activity in the osseointegrated dental means of experimental implants, the central compartimplant sulcus: A comparison of sites in edentulous and partially ment of which could be loaded with TGF-(3 or control edentulous patients. J. Periodont. Res., 24:96-105. vehicle. Despite increased osteoblastic activity and, to Artzi, Z., H. Tal, 0. Moses, and A. Kozlovsky 1993 Mucosal considera lesser extent, osteoclastic activity, the volume of traations for osseointegrated implants. J . Prosth. Dent., 70:427-432. becular bone formed after 22 days of implantation in Aufdemorte, T.B., W.C. Fox,G.R. Holt, H.S. McGuff, A.J. Ammann, and L.S. Beck 1993 An intraosseous device for studies of bonebaboon tibias was comparable to that in control imhealing. The effect of transforming growth factor beta. J . Bone plants. TGF-P1 also is a promoter of osteoclastic activJoint Surg. [Am.] 74:1153-1161. ity (Cochran et al., 1993). On the other hand, Selvig et Babbush, C.A., A. Kirsch, P.J. Mentag, and B. Hill 1987 Intramobile al. (1994) failed to demonstrate any beneficial effect on cylinder (IMZ) two-stage osteointegrated implant system with the intramobile element (IME): Part I. Its rationale and procebone regeneration, as compared to control sites, when dure for use. Int. J. Oral Maxillofac. Implants, 2:203-216. experimentally created bone defects in dog mandibles C.A., and C.A. Shimura 1993 Five-year statistical and clinwere treated with a topical application of a mixture of Babbush, ical observationswith the IMZ two-stage osteointegrated implant IGF-11, basic fibroblast growth factor (FGF), and transsystem. Int. J . Oral Maxillofac. Implants, 8:245-253. forming growth factor (TGF-(31).These apparently dis- Barth, E., C. Johansson, and T. Albrektsson 1990 Histologic comparGROWTH FACTORS AND GTR ~ . I , 422 M.A. LISTGARTEN ison of ceramic and titanium implants in cats. Int. J . Maxillofac. Implants, 5~227-231. Bauman. G.R.. M. Mills, J.W. RaDlev. and W.W. Hallmom 1992 Plaoue-inducedi n f l a k a t i o n aioind imulants. Int. J . Oral Maxil1of;lc. Implants, 7:330-337. Bauman, G.R., J.W. Rapley, W.W. Hallmon, and M. Mills 1993 The peri-implant sulcus. Int. J. Oral Maxillofac.Implants, 8:273-280. Beck. L.S.. E.P. Amento. Y. Xu. L. Dewman. W.P. Lee. T. Nrmven. and N.A. Gillett 1993 TGF-beta <induces bone closure 07 skull defects: Temporal dynamics of bone formation in defects exposed to rhTGF-beta. J . Bone Miner. Res., 8:753-761. Becker, W., C. Dahlin, B.E. Becker, U. Lekholm, D. van Steenberghe, K. Higuchi, and C. Kultje 1994 The use of e-PTFE barrier membranes for bone promotion around titanium implants placed into extraction sockets: A prospective multicenter study. Int. J. Oral Maxillofac. Implants, 9:31-40. Becker, W., S.E. Lynch, U. Lekholm, B.E. Becker, R. Caffesse, K. Donath, and R. Sanchez 1992 A comparison of e-PTFE membranes alone or in combination with platelet-derived growth factors and insulin-like growth factor-I or demineralized freezedried bone in promoting bone formation around immediate extraction socket implants. J. Periodontol., 63:929-940. Berglundh, T., J. Lindhe, I. Ericsson, C.P. Marinello, B. Liljenberg, and P. Thomsen 1991 The soft tissue barrier a t implants and teeth. Clin. Oral Implants Res., 2331-90. Berglundh, T., J. Lindhe, C. Marinello, I. Ericsson, and B. Liljenberg 1992 Soft tissue reaction to de novo plaque formation on implants and teeth. An experimental study in the dog. Clin. Oral Implants Res., 3:l-8. Bidez, M.W., and C.E. Misch 1992 Issues in bone mechanics related to oral implants. Implant Dent., 1:289-294. Birkedal-Hansen, H., W.G.I. Moore, M.K. Bodden, L.J. Windsor, B. Birkedal-Hansen, A. DeCarlo, and J.A. Engler 1993 Matrix metalloproteinases: A review. Crit. Rev. Oral Biol. Med., 4:197-250. Bosch, F.X., J.P. Ouhayoun, B.L. Bader, C. Collin, C. Grund, I. Lee, and W.W. Franke 1989 Extensive changes in cytokeratin expression patterns in pathologically affected human gingiva. Virchows Arch. B. Cell. Pathol., 58:59-77. Brimemark, P.-I., B.O. Hansson, R. Adell, U. Breine, J. Lindstrom, 0. Hallen, and A. ohman 1977 Osseointegrated implants in the treatment of the edentulous jaw. Experience from a 10-year period. Scand. J. Plast. Reconstr. Surg., 2(Suppl.):16. Brhemark, P.-I., G.A. Zarb, and T. Albrektsson 1985 Tissue Integrated Prostheses. Osseointegration in Clinical Dentistry. Quintessence, Chicago. Brunette, D.M., J. Ratkay, and B. Chehroudi 1991 Behaviour of osteoblasts on micromachined surfaces. In: The Bone-Biomaterial Interface. J.E. Davies, ed. University of Toronto Press, Toronto, pp. 170-179. Brunski, J.B. 1991 Influence of biomechanical factors a t the bonebiomaterial interface. In: The Bone-Biomaterial Interface. J.E. Davies, ed. University of Toronto Press, Toronto, pp. 391-404. Buser, D. 1987 Die Vestibulumplastik mit freien Schleimhauttransplantaten bei Implantaten im zahnlosen Unterkiefer. Schweiz. Monatschr. Zahnmed., 97:766-772. Buser, D., B.K. Schenk, S. Steinemann, J.P. Fiorellini, C.H. Fox, and H. Stich 1991a Influence of surface characteristics on bone integration of titanium implants. J . Biomed. Mater. Res., 25:889902. Buser, D., H. Stich, G. Krekeler, and A. Schroeder 1989 Faserstrukturen der periimplantaren Mukosa bei Titanimplantaten. Eine tierexperimentelle Studie am Beagel-Hund. Z. Zahnartzl. Implantol., 5: 15-23. Buser, D., K. Warrer, and T. Karring 1990a Formation of a periodontal ligament around titanium implants. J. Periodontol., 61 :597601. Buser, D., K. Warrer, T. Karring, and H. Stich 1990b Titanium implants with a true periodontal ligament: An alternative to osseointegrated implants? Int. J. Oral Maxillofac. Implants, 5:113116. Buser, D., H.P. Weber, U. Bragger, and C. Balsiger 1991b Tissue integration of one-stage IT1 implants: 3-year results of a longitudinal study with hollow-cylinder and hollow-screwimplants. Int. J. Oral Maxillofac. Implants, 6~405-412. Buser, D., H.P. Weber, and N.P. Lang 199Oc Tissue integration of non-submerged implants. 1-year results of a prospective study with 100 IT1 hollow-screw and hollow-cylinder implants. Clin. Oral Implants Res., 1:33-40. Buser, D., H.P. Weber, K. Donath, J.P. Fiorellini, D.W. Paquette, and R.C. Williams 1992 Soft tissue reactions to non-submerged un- loaded titanium implants in Beagle dogs. J. Periodontol.,63:226236. Carlsson, L., T. Riistlund, B. Albrektsson, and T. Albrektsson 1988 Removal torques for polished and rough titanium implants. Int. J . Oral Maxillofac. Implants, 3:21-24. Caudill, R. 1993 Histologic analysis of the osseointegration of endosseous implants in simulated extraction sockets with and without e-PTFE barriers. Part 11: Histomorphometric findings. J . Oral Implantol., 19:209-215. Clokie, C.M.L., and H. Warshawsky 1995 Morphologic and radioautographic studies of bone formation in relation to titanium implants using the rat tibia as a model. Int. J. Oral Maxillofac. Implants, 10:155-165. Cochran, D.L., C.A. Rouse, S.E. Lynch, and D.T. Graves 1993 Effects of platelet-derived growth factor isoforms on calcium release from neonatal mouse calvaria. Bone, 14:53-58. Cochran, D.L., J . Simpson, H.-P. Weber, and D. Buser 1994 Attachment and growth of periodontal cells on smooth and rough titanium. Int. J. Oral Maxillofac. Implants, 9:289-297. Dahlin, C., U. Lekholm, and A. Linde 1991 Membrane induced bone augmentation of titanium implants. Int. J. Periodont. Rest. Dent., 11:273-282. Dao, T.T.T., J.D. Anderson, and G.A. Zarb 1993 Is osteoporosis a risk factor for osseointegration of dental implants? Int. J. Oral Maxillofac. Implants, 8:137-144. De Lange, G., and C. De Putter 1993 Structure of the bone interface to dental implants in vivo. J. Oral Implantol., 19:123-135. Deporter, D.A., P.A. Watson, R.M. Pilliar, T.P. Howley, and J. Winslow 1988 A histological evaluation of a functional endosseous, porous-surfaced, titanium alloy dental implant system in the dog. J. Dent. Res., 67:1190-1195. Donley, T.G., and W.B. Gillette 1991 Titanium endosseous implantsoft tissue interface: A literature review. J. Periodontol., 62:153160. Ericson, L.E., B.R. Johansson, A. Rosengren, L. Sennerby, and P. Thomsen 1991 Ultrastructural investigation and analysis of the interface of retrieved metal implants. In: The Bone-Biomaterial Interface. J.E. Davies, ed. University of Toronto Press,Toronto, pp. 425-436. Ericsson, A.R., and T. Albrektsson 1983 Temperature threshold levels for heat-induced bone tissue injury: A vital-microscopic study in the rabbit. J. Prosthet. Dent., 5O:lOl-107. Ericsson, I., T. Berglundh, C. Marinello, B. Liljenberg, and J. Lindhe 1992 Long-standing plaque and gingivitis at implants and teeth in the dog. Clin. Oral Implants Res., 3:99-103. Ericsson, I., and J. Lindhe 1993 Probing depth a t implants and teeth. An experimental study in the dog. J. Clin. Periodontol., 20:623627. Ettinger, R.L., J.D. Spivey, D.-H. Han, and G.F. Koorbusch 1993 Measurement of the interface between bone and immediate endosseous implants: A pilot study in dogs. Int. J. Oral Maxillofac. Implants, 8:420-427. Gher, M.E., G. Quintero, D. Assad, E. Monaco, and A.C. Richardson 1994a Bone grafting and guided bone regeneration for immediate dental implants in humans. J . Periodontol., 65:881-891. Gher, M.E., G. Quintero, J.B. Sandifer, M. Tabacco, and A.C. Richardson 1994b Combined dental implant and guided tissue regeneration therapy in humans. Int. J. Periodont. Rest. Dent., 14: 332-347. Gotfredsen, K., L. Nimb, D. Buser, and E. Hjorting-Hansen 1993 Evaluation of guided bone regeneration around implants placed into fresh extraction sockets: An experimental study in dogs. J. Oral Maxillofac. Surg., 71:879-884. Gotfredsen, K., E. Rostrup, E. Hjorting-Hansen, K. Stoltze, and E. BudtzJorgensen 1991 Histological and histomorphometrical evaluation of tissue reactions adjacent to endosteal implants in monkeys. Clin. Oral Implants Res., 2:30-37. Gottlander, M., and T. Albrektsson 1991 Histomorphometric studies of hydroxyapatite-coated and uncoated CP titanium threaded implants in bone. Int. J. Oral Maxillofac. Implants, 6:399-404. Gottlander, M., T. Albrektsson, and L.V. Carlsson 1992 A histomorphometric study of unthreaded hydroxyapatite-coated and titanium-coated implants in rabbit bone. Int. J. Oral Maxillofac. Implants, 7:485-490. Graves, D.T., and D.L. Cochran 1994 Periodontal regeneration with polypeptide growth factors. Curr. Opin. Periodontol., 178-186. Hammerle, C.H., J. Schmid, N.P. Lang, and A.J. Olah 1995 Temporal dynamics of healing in rabbit cranial defects using guided bone regeneration. J. Oral Maxillofac. Surg., 53:167-174. Hassell, T.M. 1993 Tissues and cells of the periodontiurn. In: Peri- TISSUE RESPONSE TO DENTAL IMPLANTS odontal TissuesCtructure and Function. Periodontology 2000, Vol. 3. T.M. Hassell, ed. pp. 9-38. Munksgaard, Copenhagen. HeRi, A.F. 1993 Aspects of cell biology of the normal periodontium. In: Periodontal Tissues-Structure and Function. Periodontology 2000, Vol. 3. T.M. Hassell, ed. pp. 64-75. Munksgaard, Copenhagen. Hertel, R.C., P.A. Blijdorp, W. Kalk, and D.L. Baker 1994 Stage 2 surgical techniques in endosseousimplantation. Int. J. Oral Maxillofac. Implants, 9:273-278. James, R.A., and R.L. Schultz 1974 Hemidesmosomesand the adhesion of junctional epithelial cells to metal implants-a preliminary report. Oral Implantol., 4:294-302. Johansson, C., J. Lausmaa, M. Ask, H.A. Hansson, and T. Albrektsson 1989 Ultrastructural differences of the interface zone between bone and Ti6A14V or commercially pure titanium. J. Biomed. Eng., lr3-8. Johansson, C.B., and T. Albrektsson 1991 A removal torque and histomorphometric study of commercially pure niobium and titanium implants in rabbit bone. Clin. Oral Implants Res., 2:24-29. Johansson, C.B., L. Sennerby, and T. Albrektsson 1991 A removal torque and histomorphometric study of bone tissue reactions to commerciallypure titanium and vitallium implants. Int. J. Oral Maxillofac. Implants, 6:437-441. Jovanovic, S.A., E.B. Kenney, F.A. Carranza, Jr., and K. Donath 1993 The regenerative potential of plaque-induced peri-implant bone defects treated by a submerged membrane technique: An experimental study. Int. J . Oral Maxillofac. Implants, 8:13-18. Jovanovic, S.A., R.K. Schenk, M. Orsini, and E.B. Kenney 1995 Supracrestal bone formation around dental implants: an experimental dog study. Int. J. Oral Maxillofac. Implants, 10:23-31. Jovanovic, S.A., H. Spiekermann, and E.J. Richter 1992 Bone regeneration around titanium dental implants in dehisced defect sites: A clinical study. J . Oral Maxillofac. Implants, 7~233-245. Karring, T., S. Nyman, J. Gottlow, and L. Laurel1 1993 Development of the biologic concept of guided tissue regeneration-animal and human studies. In: Periodontal Regeneration. Periodontology 2000, Vol. 1. J.G. Caton, ed. pp. 26-35. Kiritsy, C.P., and S.E. Lynch 1993 Role of growth factors in cutaneous wound healing: A review. Crit. Rev. Oral Biol. Med., 4:726-760. Kohn, D.H. 1992 Overview of factors important in implant design. J. Oral Implantol., 18~204-219. Lang, N.P., U. Bragger, C.H. Hammerle, and F. Sutter 1994a Immediate transmucosal implants using the principle of guided tissue regeneration. I. Rationale, clinical procedures and 30-month results. Clin. Oral Implants Res., 5~154-163. Lang, N.P., U. Bragger, D. Walther, N. Beamer, and K.S. Kornman 1993 Ligature-induced pen-implant infection in cynomolgus monkeys. I. Clinical and radiographic findings. Clin. Oral Implants bs., 4:2-11. Lang, N.P., C.H. Hammerle, U. Bragger, B. Lehmann, and S.R. Nyman 1994b Guided tissue regeneration in jawbone defects prior to implant placement. Clin. Oral Implants Res., 5:92-97. Lekholm, U., R. Adell, J. Lindhe, P.-I. Brinemark, B. Eriksson, B. Rockler, A.-M. Lindvall, and T. Yoneyama 1986 Marginal tissue reactions at osseointegratedtitanium fixtures. 11. Cross-sectional retrospective study. Int. J. Oral Maxillofac. Surg., 15:53-61. Lekholm, U., W. Becker, C. Dahlin, B. Becker, K. Donath, and E. Morrison 1993 The role of early versus late removal of GTAM membranes on bone formation a t oral implants placed into immediate extraction sockets. Clin. Oral Implants Res., 4:121-129. Lemons, J.E., and M.W. Bidez 1991 Endosteal implant biomaterials and biomechanics. In: Endosteal Dental Implants. R. V. McKinney, Jr., ed. Mosby, St. Louis, pp. 27-36. Linde, A., P. Alberius, C. Dahlin, K. Bjurstam, and Y. Sundin 1993 Osteopromotion:A soft-tissue exclusion principle using a membrane for bone healing and bone neogenesis. J. Periodontol., 64: 1116-1128. Linder, L. 1989 Osseointegration of metallic implants. I. Light microscopy in the rabbit. Acta Orthop. Scand., 60:129-134. Linder, L., T. Albrektsson, P.4. Brinemark, H.-A. Hansson, B. Ivarsson, U. Jonsson, and I. LundstrGm 1983 Electron microscopic analysis of the bone-titanium interface. Acta Orthop. Scand., 54:4552. Linder, L., K. Obrant, and G. Boivin 1989 Osseointegration of metallic implants. 11. Transmission electron microscopy in the rabbit. Acta Orthop. Scand., 60:135-139. Lindhe, J., T. Berglundh, I. Ericsson, B. Liljenberg, and C. Marinello 1992 Experimental breakdown of peri-implant and periodontal tissues. A study in the beagle dog. Clin. Oral Implants Res., 3:916. 423 Listgarten, M.A. 1972a Normal development structure, physiology and repair of gingival epithelium. Oral Sci. Rev., 1:3-68. Listgarten, M.A. 1972b Ultrastructure of the dento-gingivaljunction after gingivectomy. J. Periodont. Res., 7:151-160. Listgarten, M.A. 1986 Pathogenesis of periodontitis. J. Clin. Periodontol., 13:418-425. Listgarten, M.A., D. Buser, S.G. Steinemann, K. Donath, N.P. Lang, and H.P. Weber 1992 Light and transmission electron microscopy of the intact interfaces between non-submerged titanium-coated epoxy resin implants and bone or gingiva. J. Dent. Res., 71:364371. Listgarten, M.A., and C.-H. Lai 1975 Ultrastructure of the intact interface between an endosseous epoxy resin dental implant and the host tissue. J. Biol. Buccale, 3:13-28. Listgarten, M.A., N.P. Lang, H.E. Schroeder, and A. Schroeder 1991 Periodontal tissues and their counterparts around endosseousimplants. Clin. Oral Implants Res., 2:l-19. Listgarten, M.A., S. Rosenberg, and S. Lerner 1982 Progressive replacement of epithelial attachment by a connective tissue junction after experimental periodontal surgery in rats. J. Periodontol., 53:659-670. Lum, L.B., O.R. Beirne, and D.A. Curtis 1991 Histologicevaluation of hydroxyapatite-coated versus uncoated titanium blade implants in delayed and immediately loaded applications. Int. J. Oral Maxillofac. Implants, 6:456-462. Lundgren, D., L. Sennerby, H. Falk, B. Friberg, and S. Nyman 1994 The use of a new bioresorbable barrier for guided bone regeneration in connection with implant installation. Case reports. Clin. Oral Implants Res., 5:177-184. Lynch, S. 1994 The role of growth factors in periodontal repair and regeneration. In: Periodontal Regeneration. Current Status and Directions. A.M. Polson, ed. Quintessence,Chicago, pp. 179-198. Lynch, S., D. Buser, R.A. Hernandez, H.P. Weber, H. Stich, C.H. Fox, and R.C. Williams 1991 Effects of the platelet-derived growth factor/insulin-like growth factor-I combination on bone regeneration around titanium dental implants. Results of a pilot study in beagle dogs. J . Periodontol., 62:710-717. Mackenzie, I.C. 1988 Factors influencing the stability of the gingival sulcus. In: Periodontology Today. B. Guggenheim, ed. S. Karger, Basel, pp. 41-49. Mackenzie, I.C., and N.E. Fusenig 1983 Regeneration of organized epithelial structure. J . Invest. Dermatol.. 81:1899-194s. Mackknzie, I.C., and M.W. Hill 1984 Connectbe tissue influences on patterns of epithelial architecture and keratinization in skin and oral mucosa of the adult mouse. Cell Tissue Res., 235~551-559. Mackenzie, I.C., G. Rittman, Z. Gao, I. Leigh, and E.B. Lane 1991 Patterns of cytokeratin expression in human gingival epithelium. J. Periodont. Res., 26:468-478. McCulloch, C.A.G. 1993 Basic considerations in periodontal wound healing to achieve regeneration. In: Periodontal Regeneration. Periodontology 2000, Vol. 1. J.G. Caton, ed. pp. 16-25. Munksgaard, Copenhagen. McKee, M.D., and A. Nanci 1993 Ultrastructural, cytochemical and immunocytochemicalstudies on bone and its interfaces. Cells and Materials, 3:219-243. Nanci, A., G.F. McCarthy, S. Zalzal, C.M.L. Clokie, H. Warshawsky, and M.D. McKee 1994 Tissue response to titanium implants in the rat tibia: Ultrastructural, immunocytochemical and lectincytochemical characterization of the bone-titanium interface. Cells and Materials, 4:l-30. Niki, M., G. Ito, T. Matsuda, and M. Ogino 1991 Comparative pushout data of bioactive and non-bioadive materials 6f similar rugosity. In: The Bone-Biomaterial Interface. J.E. Davies, ed. University of Toronto Press, Toronto, pp. 350-356. Nishibori, M., N.J.Betts, H. Salama, and M.A. Listgarten 1994 Shortterm healing of autogenous and allogeneic bone grafts after sinus lifting: A report of two cases. J. Periodontol., 65:958-966. Nystrom, E., K.-E. Kahnberg, and T. Albrektsson 1993 Treatment of the severely resorbed maxillae with bone graft and titanium implants: Histologic review of autopsy specimens. Int. J. Oral Maxillofac. Implants, 8:167-172. O’Neal, R.B., J.J. Sauk, and M.J. Somerman 1992 Biological requirements for material integration. J. Oral Implantol., 18:243-255. ONeal, R., H.-L. Wang, R.L. MacNeil, and M.J. Somerman 1994 Cells and materials involved in guided tissue regeneration. Curr. Opin. Periodontol., 141-156. Page, R.C., and H.E. Schroeder 1982 Periodontitis in Man and Other Animals: A Comparative Review. S. Karger, Basel. Palmer, R.M., P.D. Floyd, P.J . Palmer, B.J. Smith, C.B. Johansson, and T. Albrektsson 1994 Healing of implant dehiscence defects with and without expanded polytetrafluoroethylene membranes: 424 M.A. LISTGARTEN A controlled clinical and histological study. Clin. Oral Implants Res., 5:98-104. Parr, G.R., D.E. Steflik, and A.L. Sisk 1993 Histomorphometric and histologic observations of bone healing around immediate implants in dogs. Int. J. Oral Maxillofac. Implants, 8:534-540. Piatelli, A,, A. Ruggeri, M. Franchi, N. Romasco, and P. Risi 1993a An histologic and histomorphometric study of bone reactions to unloaded and loaded non-submerged single implants in monkeys: A pilot study. J . Oral Implantol., 19:314-319. Piatelli, A,, P. Trisi, N. Romasco, and M. Emanuelli 199313 Histologic analysis of a screw implant retrieved from man: Influence of early loading and primary stability. J. Oral Implantol., 19:303306. Pilliar, R.M. 1991 Quantitative evaluation of the effect of movement at a porous coated implant-bone interface. In: The Bone-Biomaterial Interface. J.E. Davies, ed. University of Toronto Press, Toronto, pp. 380-386. Quirynen, M., C.M.L. Bollen, H. Eyssen, and D. van Steenberghe 1994 Microbial penetration along the implant components of the Brlnemark system@.An in vitro study. Clin. Oral Implants Res., 5:239-244. Quirynen, M., and D. van Steenberghe 1993 Bacterial colonization of the internal part of two-stage implants. An in vivo study. Clin. Oral Implants Res., 5:239-244. Rahal, M.D., P.-I. Brhnemark, and D.G. Osmond 1993 Response of bone marrow to titanium implants: Osseointegration and the establishment of a bone-marrow-titanium interface in mice. Int. J . Oral Maxillofac. Implants, 8~573-579. Ripamonti, U., M. Heliotis, B. van den Heever, and A.H. Reddi 1994 Bone morphogenetic proteins induce periodontal regeneration in the baboon (Pupio ursinus). J . Periodont. Res., 29:439-445. Roberts, W.E. 1993 Fundamental principles of bone physiology, metabolism and loading. In: Osseointegration in Oral Rehabilitation. I. Naert, D. van Steenberghe, P. Worthington, eds. Quintessence, London, pp. 157-170. Roberts, W.E., K.J. Marshall, and P.G. Mozsary 1989 Rigid endosseous implant utilized as anchorage to protract molars and close an atrophic extraction site. Angle Orthod., 60:135-152. Roberts, E.W., L.C. Poon, and R.K. Smith 1986 Interface histology of rigid endosseous implants. Oral Implantol., 12:406-416. Roberts, W.E., K.E. Simmons, L.P. Garetto, and R.A. DeCastro 1992 Bone physiology and metabolism in dental implantology: risk factors for osteoporosis and other metabolic bone diseases. Implant. Dent., 1:11-21. Roberts, W.E., R.K. Smith, Y. Zilberman, P.G. Mozsary, and R.S. Smith 1984 Osseous adaptation to continuous loading of rigid endosseous implants. Am. J . Orthod., 86:95-111. Rominger J.W., and R.G. Triplett 1994 The use of guided tissue regeneration to improve implant osseointegration. J . Oral Maxillofac. Surg., 52:106-112. Sailer, H.F., and E. Kolb 1994 Application of purified bone morphogenetic protein (BMP) on cranio-maxillo-facial surgery. BMP in compromised surgical reconstructions using titanium implants. J. Craniomaxillofac. Surg., 22:2-11. Schenk, R.K., D. Buser, W.R. Hardwick, and C. Dahlin 1994 Healing pattern of bone regeneration in membrane-protected defects: A histological study in the canine mandible. Int. J . Oral Maxillofac. Implants, 9:13-29. Schou, S., P. Holmstrup, E. Hjorting-Hansen, and N.P. Lang 1992 Plaque-induced marginal tissue reactions of osseointegrated oral implants: A review of the literature. Clin. Oral Implants Res., 3:149-161. Schou, S., P. Holmstrup, K. Stoltze, E. Hjorting-Hansen, and K.S. Kornman 1993 Ligature-induced marginal inflammation around osseointegrated implants and ankylosed teeth. Clin. Oral Implants Res., 4:12-22. Schroeder, A,, 0. Pohler, and F. Sutter 1976 Gewebereaktion auf ein Titan-Hohlzvlinderimplantatmit Titan-Spritzschichtoberflache. Schweiz. M6natszeitsihr. Zahnheilk., 86:?13-727. Schroeder, A., F. Sutter, and G. Krekeler 1991 Oral Implantology. Georg Thieme Verlag, Stuttgart. Schroeder, A., E. van der Zypen, H. Stich, and F. Sutter 1981 The reaction of bone, connective tissue and epithelium to endosteal implants with sprayed titanium surfaces. J. Maxillo. Fac. Surg., 9: 15-25. Schroeder, H.E. 1986 The periodontium. In: Handbook of Microscopic Anatomy, Vol 5. Springer-Verlag, Berlin. Schroeder, H.E. 1991 Oral Structural Biology. George Thieme Verlag, Stuttgart. Schroeder, H.E., and M.A. Listgarten 1977 Fine structure of the developing epithelial attachment of human teeth. In: Monographs in Developmental Biology, 2nd ed. A. Wolsky, ed. S. Karger, Basel. Schiipbach, P., M. Hiirzeler, and U. Grunder 1994 Implant-tissue interfaces following treatment of periimplantitis using guided tissue regeneration. A light and electron microscopic study. Clin. Oral Implants Res., 5:55-65. Seibert, J., and S. Nyman 1990 Localized ridge augmentation in dogs: A pilot study using membranes and hydroxyapatite. J . Periodontol., 61:157-165. Selvig, K.A., U.M. Wikesjo, G.C. Bogle, and R.D. Finkelman 1994 Imuaired early bone formation in Deriodontal fenestration defects in dogs following application of ihsulin-like factor (11). Basic fibroblast growth factor and transforming growth factor beta 1.J. Clin. Periodontol., 21:380-385. Sennerby, L., L.E. Ericson, P. Thomsen, U.Lekholm, and P. Astrand 1991 Structure of the bone-titanium interface in retrieved clinical oral implants. Clin. Oral Implants Res., 2~103-111. Sennerby, L., P. Thomsen, and L.E. Ericson 1992 A morphometric and biomechanic comparison of titanium implants inserted in rabbit cortical and cancellous bone. Int. J. Oral Maxillofac. Implants, 7:62-71. Shanaman, R.H. 1994 A retrospective study of 237 sites treated consecutively with guided tissue regeneration. Int. J. Periodont. Rest. Dent., 14:293-302. Sharf, D.R., and D.P. Tarnow 1993 Success rates of osseointegration for implants placed under sterile versus clean conditions. J . Periodontol., 64:954-956. Skalak, R. 1983 Biomechanical considerations in osseointegrated prostheses. J. Prosthet. Dent., 49:843-848. Smith, D.C. 1993 Dental implants: Materials and design considerations. Int. J. Prosthodont., 6:106-117. Smith, K.G., C.D. Franklin, R. van Noort, and D.J. Lamb 1992 Tissue response to the implantation of two new machinable calcium phosphate ceramics. Int. J . Oral Maxillofac. Implants, 7:395400. Steflik, D.E., P.J. Hanes, A.L. Sisk, G.R. Parr, M.J. Song, F.T. Lake, and R.V. McKinney 1992a Transmission electron microscopic and high voltage electron microscopic observations of the bone and osteocyte activity adjacent to unloaded dental implants placed in dogs. J. Periodontol., 63:443-452. Steflik, D.E., G.R. Parr, A.L. Sisk, P.J. Hanes, and F.T. Lake 199213 Electron microscopy of bone response to titanium cylindrical screw-type endosseous dental implants. Int. J. Oral Maxillofac. Implants, 7:497-507. Steflik. D.E.. A.L. Sisk, G.R. Parr. L.K. Gardner. P.J. Hanes, F.T. Lake, D.J. Berkery, and P. Brewer 1993a Osteogenesis at the dental implant interface: high-voltage electron microscopic and conventional transmission electron microscopic observations. J. Biomed. Mater. Res., 27:791-800. Steflik, D.E., A.L. Sisk, G.A. Parr, F.T. Lake, and P.J. Hanes 1993b Experimental studies of the implant-tissue interface. J. Oral Implantol., 19:90-94. Teerlinck, J., M. Quirynen, P. Darius, and D. van Steenberghe 1991 Periotest: An objective clinical diagnosis of bone apposition toward implants. Int. J. Oral Maxillofac. Implants, 6:55-61. ten Bruggenkate, C.M., G. Krekeler, W.A.M. van der Kwast, and H.S. Oosterbeek 1990 Palatal mucosal grafts for oral implant devices. In: Clinical and Radiological Aspects of Oral Implants, With Special Emphasis on the I.T.I. Hollow Cylinder Implant. C.M. ten Bruggengate, ed. Thesis, Free University of Amsterdam. Pasmans, Gravenhage, pp. 49-58. Teranobu, O., I. Naito, N. Takahashi, M. Fujiwara, K. Iwata, M. Umeda, K. Shimada, K. Kurioka, H. Kawamoto, K. Shimada, and T. Kawai 1989 The influence of sintering condition and surface shape of hydroxyapatite ceramics for the osteoconductivity. In: Oral Implantology and Biomaterials, Progress in Biomedical Engineering, Vol. 7. H. Kawhara, ed. Elsevier, New York, pp. 233238. Wang, H.L., T.D. Pappert, W.A. Castelli, D.J. Chiego, Jr., Y. Shyr, and B.A. Smith 1994 The effect of platelet-derived growth factor on the cellular response of the periodontium: An autoradiographic study on dogs. J. Periodontol., 65:429-436. Warrer, K., K. Gotfredsen, E. Hjorting-Hansen, and T. Karring 1991 Guided tissue regeneration ensures osseointegration of dental implants placed into extraction sockets. An experimental study in monkeys. Clin. Oral Implants Res., 2:166-171. Warrer, K., T. Karring, and K. Gotfredsen 1993 Periodontal ligament formation around different types of dental titanium implants. I. The self-tapping screw type implant system. J . Periodontol., 64: 29-34. Wehrbein, H., and P. Diedrich 1993 Endosseous titanium implants TISSUE RESPONSE TO DENTAL IMPLANTS during and after orthodontic load-an experimental study in the dog. Clin. Oral Implants Res., 4:76-82. Weinlaender, M., E.B. Kenney, V. Lekovik, J. Beumer, P.K. Moy, and S. Lewis 1992Histomorphometry of bone apposition around three types of endosseous dental implants. Int. J . Oral Maxillofac. Implants, 7:491-496. Weiss, C.M. 1986 Tissue integration of dental endosseous implants: Description and comparative analysis of the fibrosseous integration and osseous integration systems. J . Oral Implantol., 12:169214. Wennstrom, J.L., F.Bengazi, and U. Lekholm 1994 The influence of the masticatory mucosa on the peri-implant so& tissue condition. Clin. Oral Implants Res., 5:l-8. 425 Wilke, H.J., L. Claes, and S. Steinemann 1990 The influence of various titanium surfaces on the interface shear strength between implants and bone. In: Clinical Implant Materials, Advances in Biomaterials, Vol. 9.Elsevier, New York, pp. 309-314. Zablotsky, M., R.Meffert, R. Caudill, and G. Evans 1991 Histological and clinical comparisons of guided tissue regeneration on dehisced hydroxylapatite-coated and titanium endosseous implant surfaces: A pilot study. Int. J. Oral Maxillofac. Implants, 6.294303. Zablotsky, M.H. 1992 Hydroxyapatite coatings in implant dentistry. Implant Dent., 1.253-257.