Biology and Principles of Periodontal Wound Healing/Regeneration

Biology and principles of periodontal wound healing/regeneration

GIUSEPPE POLIMENI,ANDREAS V. XIROPAIDIS &ULF M. E. WIKESJO¨

The native periodontium includes cementum, a Predictability of outcomes following surgical pro- functionally oriented periodontal ligament, alveolar cedures is of fundamental importance in medicine. bone and gingiva. Pathologic and/or traumatic events As periodontal-regenerative procedures are time may lead to the loss or damage of this anatomical consuming and financially demanding, there is structure. Since the 1970s, a number of procedures increasing interest by clinicians to learn of factors have been investigated in an attempt to restore such that may influence the clinical outcome following lost tissues. Numerous clinical trials have shown periodontal reconstructive surgery in order to pro- positive outcomes for various reconstructive surgical vide the best possible service to patients. This goal protocols. Reduced probing depths, clinical attach- can only be achieved if biological aspects of wound ment gain, and radiographic bone fill have been healing and regeneration are taken into considera- reported extensively for intrabony and furcation de- tion. The objectives of the present article are to fects following scaling and root planing, open flap provide an overview of wound healing following debridement, autogenous bone grafting, implanta- periodontal surgical procedures, to discuss the basic tion of biomaterials including bone derivatives and principles of periodontal regeneration, and to illus- bone substitutes, guided-tissue regeneration (GTR) trate the factors that influence this process. procedures, and implantation of biologic factors, including enamel matrix proteins. Histological studies have shown that various Wound healing surgical periodontal procedures can lead to different patterns of healing. Healing by formation of a The healing of wounds in nonoral sites has been long junctional epithelium (epithelial attachment) is studied in great detail (Fig. 1) (5). The general prin- characterized by a thin epithelium extending apically ciples of healing, and the cellular and molecular interposed between the root surface and the gingival events observed in nonoral sites, also apply to the connective tissue (4, 23). Connective tissue repair healing processes that take place following perio- (new attachment) is represented by collagen fibers dontal surgery. Traumatic injury causes capillary oriented parallel or perpendicular to a root surface damage and hemorrhage and, as a result, a blood clot previously exposed to periodontal disease or other- is formed. The formation of a clot is the immediate wise deprived of its periodontal attachment. In con- response to any trauma. The clot has two functions: it trast, periodontal regeneration is characterized by temporarily protects the denuded tissues; and it de novo formation of cementum, a functionally ori- serves as a provisional matrix for cell migration (5). ented periodontal ligament, alveolar bone, and gin- The blood clot consists of all cellular components of giva (restitutio ad integrum). Nevertheless, it would blood (including red and white blood cells and be naive to expect these to occur as distinctly sep- platelets) in a matrix of fibrin, plasma fibronectin, arate biologic outcomes following reconstruction of vitronectin, and thrombosporin (24). the periodontal attachment. For example, periodon- Clot formation is followed by an early stage of tal regeneration should be expected to include ele- inflammation. Within hours of injury, inflammatory ments of a new, as well as an epithelial, attachment. cells (predominantly neutrophils and monocytes)

30 Biology and principles of periodontal wound healing/regeneration

ally the breach in the epithelium is sealed. The epithelial cells in normal gingival tissues use surface receptors, known as integrins, to bind to laminin in the

Early Late basal lamina. In order to initiate migration, the kera- phase phase tinocytes dissolve this attachment to start expressing integrins suitable for the wound environment (1). Maturation of the granulation tissue will lead to the regeneration or repair (scar formation) of the injured tissues. Whether the damaged tissues heal by regeneration or repair depends upon two crucial factors: the availability of cell type(s) needed; and the presence or absence of cues and signals necessary to recruit and stimulate these cells (8). This summary represents Fig. 1. Phases of wound healing (epidermal incisional an oversimplified explanation of the wound-healing wounds), including an early (within hours) and a late (within days) phase of inflammation dominated by poly- process. The stages described may overlap consider- morphonuclear neutrophils and macrophages, respect- ably and the time needed for completion of each stage ively. The magnitude of wound contraction parallels the may vary, depending on local and systemic factors. phase of granulation tissue formation. Collagen accumu- lation is first observed during the phase of granulation tissue formation, continuing through the phase of matrix formation and remodeling. Redrawn with permission Periodontal wound healing from Dr Richard AF Clark. A more complex situation presents itself when a populate the clot. These cells cleanse the wound of mucoperiosteal flap is apposed to an instrumented bacteria and necrotic tissue through phagocytosis root surface deprived of its periodontal attachment. and release of enzymes and toxic oxygen products. In this case, the wound margins are not two opposing Within 3 days, the inflammatory reaction moves into vascular gingival margins but comprise the rigid its late phase. Macrophages migrate into the wound nonvascular mineralized tooth surface, on the one area and, in addition to wound debridement, secrete hand, and the connective tissue and epithelium of polypeptide mediators targeting cells involved in the the gingival flap, on the other. The periodontal wound-healing process (5, 45). wound also includes tissue resources from the The macrophage plays an important role in the alveolar bone and the periodontal ligament. Clot formation of granulation tissue. Growth factors and formation at the interface between the tooth and a cytokines secreted by macrophages are involved in gingival flap is initiated as blood elements are im- the proliferation and migration of fibroblasts, posed onto the root surface during surgery and at endothelial cells, and smooth muscle cells into the wound closure in a seemingly random manner, much wound area. The cell-rich granulation tissue next like a Jackson Pollock dripped and poured style undergoes maturation and remodeling. Fibroblasts canvas (http://www.nga.gov/feature/pollock/pollock responsible for the replacement of the provisional home.html [accessed 16 February 2006]). This extracellular matrix produce a new collagen-rich represents the very first healing event at the tooth– matrix. Approximately 1 week following wounding, gingival flap interface (i.e. the absorption and adhe- and once the collagen matrix has been synthesized, sion of plasma proteins onto the root surface) (Fig. 2) some fibroblasts undergo transformation into myo- (49). Within minutes, a fibrin clot attached to the root fibroblasts and express a-smooth muscle actin. This surface is developed. Within hours, one may observe transformation and synthesis is responsible for the early phase of inflammation as inflammatory wound contraction. Endothelial cells, responsible for cells, predominantly neutrophils and monocytes, angiogenesis, migrate into the provisional wound accumulate on the root surface, and within 3 days matrix to form vascular tubes and loops, and as the the late phase of inflammation dominates the healing provisional matrix matures, the endothelial cells un- picture as macrophages migrate into the wound fol- dergo programmed cell death (apoptosis) and the lowed by the formation of granulation tissue. At number of vascular units is reduced (1, 24). 7 days, a connective tissue attachment may be seen Epithelization of the wound is initiated within hours at the root surface; however, areas of the fibrin clot in of injury. Epithelial cells from the basal layer prolif- various stages of maturation may also be observed, erate and migrate through the fibrin clot and eventu- depending on wound volume and tissue resources.

31 Polimeni et al.

Fig. 2. Early healing events at the tooth–gingival flap the dentin surface (photomicrograph ·450). (D) Late interface. (A) Red blood cells (RBC) in a granular preci- phase of inflammation, 3 days following wound closure, pitate adhering to the dentin surface, shown immediately showing macrophages lining the dentin surface (trans- (10 min) upon wound closure; an artifactual split (arrows) mission electron micrograph, ·10,500). (E) Granulation between the red blood cells and the dentin-adhering tissue formation, including fibroblasts in the maturing precipitate verifies the absorption/adhesion of blood ele- fibrin clot at the dentin surface (photomicrograph, ·450). ments to the dentin surface (transmission electron (F) Cell-rich connective tissue closely adapted to the micrograph, ·4000). (B) Red blood cells in a fibrin network dentin surface at 7 days following wound closure (pho- in close proximity to the root surface observed within 1 h tomicrograph, ·450). D, Dentin/root surface; F, fibrin; FB, of wound closure (photomicrograph ·450). (C) Early fibroblast; M, macrophage; P, plasma precipitate; PMN, phase of inflammation: RBC aggregates are loosely inter- polymorphonuclear neutrophils; RBC, red blood cells;. For spersed in an organized fibrin network at 6 h. The fibrin detail see: Wikesjo¨ et al. 1991 (49). These figures are clot appears to be attached to the dentin surface. copyrighted by and modified with permission from the Numerous polymorphonuclear cells are observed lining American Academy of Periodontology.

32 Biology and principles of periodontal wound healing/regeneration

The studies described above investigated the attachment following periodontal surgery is deter- absorption, adhesion, and structural maturation of mined by the cells repopulating the root surface. the fibrin clot in periodontal wound healing, but have Karring, Nyman et al. (14) corroborated these not taken into consideration the functional integrity concepts in a series of experiments. They asked: ÔCan of the tooth–gingival flap interface. Only a few a new connective tissue attachment be established to experimental studies have as such evaluated the a root surface previously exposed to the oral envi- functional integrity of a maturing periodontal wound. ronment and implanted into bone?Õ (13). In a dog Hiatt et al. (11) examined the tensile strength of the model, they extracted and crown-resected perio- tooth–gingival flap interface following reconstructive dontitis-affected teeth, scaled and root planed the surgery of relatively small surgical dehiscence defects portion of the root that was affected with periodon- over the maxillary canine teeth in the dog. They titis, and implanted the roots into cavities created in found that the tensile strength increases from 200 g the alveolar bone. While connective tissue repair did at 3 days postsurgery to 340 g at 5–7 days postsur- not occur at the periodontitis-affected portion of the gery, and to >1700 g at 2 weeks postsurgery. In other roots, the portion where the periodontal ligament words, they found that a relatively limited perio- was preserved showed an attachment with func- dontal wound might not reach functional integrity tionally oriented periodontal fibers. until 2 weeks postsurgery. These data suggest that A second study answered the question: ÔCan a new wound integrity during the early healing phase rests connective tissue attachment establish to a perio- primarily on the stabilization of the gingival flaps dontitis-affected root implanted into gingival con- offered by suturing. In consequence, the choice of nective tissue?Õ (26). Using the dog model, they suture material, placement and removal, for perio- implanted teeth oriented in such a way that one side dontal surgical procedures aimed at regeneration of the tooth faced alveolar bone while the other faced should be dictated by such observations as should the gingival connective tissue. Similarly to the previ- postoperative protocols be instituted aimed at pro- ous study, no connective tissue attachment was tecting the surgical site from trauma from oral hy- observed at the periodontitis-affected portion of giene procedures, and plaque colonization and the roots. Conversely, the half of the root where infection (21, 37). periodontal ligament was preserved showed a con- As mentioned above, the regeneration of lost tis- nective tissue attachment. sues depends on the availability of the cell type(s) They next asked: ÔCan a new connective tissue needed and the presence or absence of cues and attachment establish to root surface deprived of its signals necessary to recruit and stimulate these cells. periodontal attachment giving preference to cells The extracellular matrix regulates how cells respond from the periodontal ligament?Õ (27). Periodontal to these signals. Stem cells responsible for regener- fenestration defects were created at the maxillary ation of the periodontal tissues reside within the lateral incisors and mandibular canines in nonhu- periodontal ligament (8). The innate regenerative man primates. A Millipore filter was placed to cover potential of the periodontium has been investigated the fenestration defects with the aim of preventing extensively and clearly appears to be dependent on gingival connective tissue from contacting the root wound management (see below). Current research surface. New cementum with a functionally oriented focuses on identifying biologic factors that favor periodontal ligament was observed within a 6-month migration and proliferation of periodontal tissues healing interval. Using the same rationale and tech- and to use those to alter the microenvironment of the nology, they subsequently provided the first evidence wound, favoring unimpeded healing and regener- that periodontal regeneration can be obtained at a ation of the periodontium. periodontitis-affected tooth in a human (28). This series, and several associated studies by Karring, Nyman et al., established that cells from Biological factors at the base of periodontal ligament have the capacity to regenerate periodontal regeneration the periodontal attachment, while the alveolar bone and gingival connective tissue do not possess this Melcher (25) postulated biological concepts at the ability (14). As postulated by Melcher, these findings base of periodontal regeneration. Accordingly, suggest that if preference is provided to cells origin- periodontal structures are subdivided in four ating from the periodontal ligament, periodontal compartments (gingival corium, periodontal liga- regeneration may consistently occur (25). It also ment, cementum, and bone) and the nature of the new appears from these studies that occlusion of cells

33 Polimeni et al. originating from the gingiva by means of tissue bar- its maturation into a connective tissue attachment. riers, also known as GTR techniques, is of paramount Collectively, these studies all appear to support the importance in achieving periodontal regeneration. vital importance of unimpeded absorption, adhesion Although Karring, Nyman et al. elegantly pioneered and maturation of the fibrin clot for formation of a and elucidated biological concepts at the base of connective tissue attachment over a long junctional periodontal regeneration, a review of additional epithelium, as discussed above. studies investigating wound-healing dynamics and Our laboratories have developed and characterized maturation in periodontal defects suggest that addi- a preclinical model, designated as the Critical-size tional factors play a role. Early observations by Supraalveolar Periodontal Defect Model (Figs 3 and 4) Linghorne & O’Connell (22) suggest that a lack of (17, 18, 44, 50). This animal model does not sponta- mechanical stability of the wound is the main factor neously regenerate following reconstructive surgery in the formation of a long junctional epithelium. without adjunctive measures. In addition, it allows Hiatt et al. (11) demonstrated a fundamental role of clinically relevant periodontal regeneration, induced the root surface-adhering fibrin clot as a preventive or supported by implanted biologics, biomaterials, or measure to apical migration of the gingival epithe- devices over that in a surgical control. We initially lium. Polson & Proye (34), using a monkey model, used this model to evaluate the significance of fibrin pointed to the importance of the unimpeded ab- clot absorption, adhesion and maturation to the root sorption, adhesion and maturation of a fibrin clot in surface. In a first study, root surfaces were coated periodontal wound healing. In brief, teeth were re- with heparin, to potentially interfere with fibrin clot implanted after root planing of the coronal third of formation/absorption/adhesion. It was shown that the root (control) or root planing followed by root root surfaces coated with heparin exhibited forma- surface demineralization with citric acid. While tion of an epithelial attachment (long junctional healing by long junctional epithelium occurred in the epithelium) whereas at control sites conditioned with controls, the root surface-demineralized teeth ex- saline, the epithelium was arrested at or immediately hibited a Ôfibrin linkageÕ maturing into a connective apical to the cemento–enamel junction (Fig. 5) (48). tissue attachment. Apparently, root surface demin- Apparently, the heparin coating compromised local eralization provided a stable anchorage of the fibrin fibrin clot formation/absorption/adhesion, either by clot over that of the root planed-only teeth, allowing compromising the clotting cascade, or by some

Fig. 3. The critical-size, supraalveolar periodontal defect defect and the yellow arrowheads represent the cemento- model. The alveolar bone and periodontal attachment, enamel junction. The defect height (vertical green ar- including the cementum, are surgically reduced circum- row), bone regeneration height (vertical yellow arrow), ferentially around the third and fourth mandibular pre- defect area (blue lines) delineated by an expanded molar teeth to a level 5–6 mm from the cemento–enamel polytetrafluoroethylene (ePTFE) membrane in this junction. The first molar is reduced to the level of the example (membrane height: vertical blue arrow), and reduced alveolar bone and the first and second premo- bone regeneration area (orange lines) are shown. The lars are extracted. Experimental treatments are applied white irregular ÔghostÕ structures within the wound area immediately upon defect induction. Wound closure for and regenerated alveolar bone represent a bone bioma- primary intention healing may be transgingival, leaving terial evaluated in this example. For detail see: Wikesjo¨ the tooth structure intact, or submerged following and Nilveús 1991 (44); Wikesjo¨ et al. 1994 (50); Koo et al. reduction of the clinical crowns. Examples of histometric 2004 (18); Koo et al. 2004 (17). These figures are copy- parameters evaluated in the critical-size, supraalveolar righted by and modified with permission from Blackwell periodontal defect model are shown: The green line and Munksgaard. arrowheads represent the base of the surgically created

34 Biology and principles of periodontal wound healing/regeneration

Fig. 4. The critical-size, supraalveolar periodontal defect or a cementum-like tissue extending from the apical model. The photomicrograph shows a representative extension of the defect. Bone regeneration is limited to section of a sham-surgery control following transgingival <25% of the defect height following a 4- or 8-week wound closure and a 4-week healing interval. The green healing interval, indicating that in control sites bone arrowhead identifies the apical extension of the defect regeneration is exhausted within 4 weeks. These char- (see Fig. 3) and the red arrowhead delineates the extent acteristics provide a discriminating critical-size model for of alveolar regeneration. The schematic illustration evaluation of the clinical potential of implantable/ shows healing, expressed as a percentage of the defect injectable devices, biomaterials, biologics, and cell con- height, in the critical-size, supraalveolar periodontal de- structs, with or without root surface biomodifications. fect model following a 4-week healing interval and Substantial regeneration in this discriminating model transgingival wound closure, and following an 8-week warrants clinical pursuit. Limited regeneration appears healing interval and submerged wound closure. Note that less deserving. For detail see: Wikesjo¨ and Nilveús 1991 the epithelium is arrested at or immediately below the (44); Wikesjo¨ et al. 1994 (50); Koo et al. 2004 (18); Koo cemento–enamel junction in sham-surgery control sites. et al. 2004 (17). These figures are copyrighted by and There is limited, if any, regeneration of the periodontal modified with permission from the American Academy of attachment, as evaluated by regeneration of cementum Periodontology. nonspecific surface action, or by a combination of ing the compromised fragile fibrin clot from wound- these effects. This single experimental manipulation rupturing forces acting on the gingival margins. apparently prevented the maturation of the fibrin clot These studies suggest that provided adequate wound into a connective tissue attachment but resulted in stability, periodontal wound healing may result in the epithelial migration and proliferation along the root formation of a connective tissue attachment rather surface, probably as a consequence of exposure of than an epithelial attachment (long junctional epi- the compromised fibrin clot to wound-rupturing thelium), which in turn should be considered a forces acting on the gingival margins. In contrast, consequence of wound failure. In perspective, when a polylactic acid implant or expanded poly- modification of the root surface by application of tetrafluoroethylene (ePTFE) membranes supported etching and chelating agents may enhance fibrin clot the gingival flaps in heparin-coated defects, the adhesion (2, 3) and promote a connective tissue tooth–gingival flap interface healed by formation of attachment (34, 47). In contrast, conditioning the connective tissue rather than by forming an epithelial root surface with protein constructs may compro- attachment (Fig. 6) (9, 43). In wound sites stabilized mise fibrin clot adhesion and, consequently, perio- by means of the polylactic acid implant or ePTFE dontal regeneration (3, 46). membranes, the epithelium was arrested coronally at Altogether, the evidence suggests that wound sta- some distance from the implant or membrane, in bility is essential for the establishment of a new con- itself indicating that wound stability, and not tissue nective tissue attachment to a root surface deprived of occlusion, played a fundamental role in the outcomes its periodontal attachment, and that tissue resources of healing. Apparently the polylactic acid implant and originating from the periodontal ligament represent a the ePTFE membrane stabilized the wound, protect- single source for periodontal regeneration, providing

35 Polimeni et al.

Fig. 6. Critical-size, 5-mm, supraalveolar periodontal de- Fig. 5. Critical-size, 5-mm, supraalveolar periodontal fect including coating of the root surfaces with a heparin defect including coating the root surfaces with a heparin solution immediately prior to wound closure with the intent solution immediately prior to wound closure with the intent to interfere with local fibrin clot formation/absorption/ to interfere with local fibrin clot formation/absorption/ adhesion. The clinical series shows the defect, the root adhesion. The clinical series shows the defect, the root surfaces isolated with a rubber dam for the heparin appli- surfaces isolated with a rubber dam for the heparin application, placement of an expanded polytetrafluoroethylene cation, transgingival wound closure, and healing at (ePTFE) membrane with the intent to stabilize the wound, 4 weeks. The left photomicrographs show sites that have and transgingival wound closure. Control defects were received the heparin coating. The green arrowheads indi- coated with heparin but did not receive ePTFE membranes. cate the base of the defects and the blue arrowhead the The photomicrographs show the epithelium arrested at the apical termination of the epithelial attachment (long cemento–enamel junction at some distance from the cor- junctional epithelium) formed in these sites. Apparently, onal extension of the ePTFE membrane (blue arrow) in the heparin coating compromised the fibrin clot in the heparin-coated defects. The controls (not shown) exhibited tooth–gingival flap interface to such an extent that allowed formation of an epithelial attachment (long junctional apical migration and proliferation of cells from the gingival epithelium). Similar observations were made in heparin- epithelium rather than maturation into a connective tissue coated defects implanted with a polylactic acid block bio- attachment. In contrast, the epithelium is arrested at the material. Collectively, these observations suggest that the cemento–enamel junction in control sites (right) treated implanted device or biomaterial provided some stability to with saline, leaving the entire denuded root surface with a the heparin-compromised tooth–gingival flap interface, new connective tissue attachment. This singular manipu- allowing the provisionary matrix (i.e. the fibrin clot) to lation aimed at interfering with coagulum formation/ mature into a connective tissue attachment rather than absorption/adhesion points to the critical importance of migration and proliferation of cells from the gingival epi- the provisionary matrix of the fibrin clot in periodontal thelium, resulting in formation of an epithelial attachment. wound healing and ultimately periodontal regeneration. Healing interval 4 weeks. For detail see: Haney et al. 1993 Healing interval 4 weeks. For detail see: Wikesjo¨ et al. 1991 (9); Wikesjo¨ and Nilveús 1990 (43). These figures are copy- (48). These figures are copyrighted by and modified with righted by and modified with permission from the Amer- permission from Blackwell Munksgaard. ican Academy of Periodontology. cells with the ability to differentiate into cemento- Cortellini & Tonetti (6) suggested decision trees, blasts, fibroblasts, and osteoblasts. Detachment of the along these lines, to provide clinicians with direction maturating fibrin clot from the root surface owing to a in their treatment of periodontal intrabony defects. lack of wound stabilization will inexorably compro- Again, patient factors and defect morphology appear mise periodontal wound healing, ultimately jeopard- to be crucial for the direction of therapy. In the fol- izing the regenerative process. lowing we use biologic observations in the Critical- size Supraalveolar Periodontal Defect Model to elucidate factors, including wound maturation, tissue Clinical and biologic variables occlusion, primary intention healing, wound failure affecting periodontal regeneration and membrane exposure, defect characteristics, space provision, and innate regenerative potential, Kornman & Robertson (20) classified factors that may that clinicians may need to consider in the regener- influence the successful management of periodontal ative treatment of periodontal defects. osseous defects. Their classification includes: • Bacterial contamination. Wound maturation • Innate wound-healing potential. • Local site characteristics. Haney et al. (9) evaluated periodontal wound healing • Surgical procedure/technique. associated with GTR membranes in supraalveolar

36 Biology and principles of periodontal wound healing/regeneration

Fig. 7. Critical-size, 5-mm, supraalveolar periodontal shows a membrane collapsed or compressed onto the defect, including coating of the root surfaces with a root surface, obstructing any regeneration of periodontal heparin solution immediately prior to wound closure, structures. There was a significant correlation between with the intent to interfere with local fibrin clot forma- the space provided by the membrane and the newly tion/absorption/adhesion. The clinical series shows the formed alveolar bone (r ¼ 0.997; P ¼ 0.002). The right defect, the root surfaces isolated with a rubber dam for photomicrograph (yellow arrow) shows an experimental heparin application, placement of an expanded poly- site with wound failure, membrane exposure, infection, tetrafluoroethylene (ePTFE) membrane with the intent to inflammation and necrosis. The green arrowheads stabilize the wound, and transgingival wound closure. delineate the apical extension of the defects. Healing Control defects were coated with heparin but did not interval 4 weeks. For detail see: Haney et al. 1993 (9). receive ePTFE membranes. The left photomicrograph These figures are copyrighted by and modified with shows a site where the ePTFE membrane allows a space permission from the American Academy of Period- at the root surface, resulting in complete fill with newly ontology. formed alveolar bone. The center photomicrograph periodontal defects and observed that most of the the concept of tissue occlusion has received limited space adjacent to the teeth underneath the mem- attention. Karaki et al. (12) evaluated bone formation branes filled with alveolar bone within a 4-week in periodontal sites using surgically created con- healing interval (Fig. 7) (9). However, there was lim- tralateral horizontal periodontal defects in the ited, if any, appreciable regeneration of cementum mandibular premolar region in dogs. A tissue- and a functionally oriented periodontal ligament, as expanding gold mesh was applied on one side, while evaluated by incandescent light microscopy, also the contralateral side served as a sham-surgery con- observed in subsequent studies using a 4-week trol. Compared with the surgical control, bone for- healing interval (19, 54). In contrast, evaluations of mation was enhanced in defects receiving the gold periodontal regeneration in supraalveolar periodon- mesh. Evidently, osteogenesis in a periodontal envi- tal defects using incandescent light microscopy and ronment may proceed in the presence of space pro- healing intervals of 8 or 24 weeks demonstrated that vision without strict occlusion of the gingival the observed bone formation is accompanied by the connective tissues. A concept of regeneration util- regeneration of cementum and a functionally orien- izing space provision without connective tissue ted periodontal ligament (Fig 8–10) (15, 38, 51–53). As occlusion emerges from this observation. Thus, a experimental conditions were similar among these study was initiated to evaluate the possibility of studies, these observations point to the possibility of periodontal regeneration without gingival tissue a delayed structural maturation of the periodontal occlusion (52). Structurally reinforced, spaceprovi- attachment compared with that of the alveolar bone ding, macroporous ePTFE membranes were sur- following regenerative procedures. gically implanted into supraalveolar periodontal defects and compared with occlusive membranes Tissue occlusion (Fig. 10). These observations clearly demonstrate that tissue occlusion is not an absolute requirement for Design criteria for GTR membranes include bio- periodontal regeneration, as sites receiving the por- compatibility, cell occlusion, space maintenance, ous membrane showed significant regeneration of tissue integration, and ease of use (10, 36). Although cementum, a functionally oriented periodontal liga- biocompatibility, space maintenance, tissue integra- ment and alveolar bone similar to that observed at tion, and ease of use have been evaluated extensively, sites receiving the occlusive membrane. There were,

37 Polimeni et al.

Fig. 8. Critical-size, 5-mm, supraalveolar periodontal gery control also with minimal regeneration, the muco- defect implanted with an occlusive, space-providing gingival flap being collapsed or compressed onto the expanded polytetrafluoroethylene (ePTFE) membrane. root. Finally, the fourth photomicrograph with a yellow The green arrow points to newly regenerated bone arrow shows a site where the membrane has been ex- reaching from the apical aspect of the defect to the ce- posed to the oral cavity, resulting in infection and nec- mento–enamel junction, the ePTFE membrane provides rosis without any regeneration of periodontal tissues. a suitable space for periodontal regeneration, and the This study points to the critical importance of primary green arrowheads delineate the apical aspect of the intention wound healing and unobstructed space provi- supraalveolar periodontal defect. The second photomi- sion for periodontal regeneration. Healing interval crograph shows the membrane collapsed or compressed 8 weeks. For detail see: Sigurdsson et al. 1994 (38). These onto the root, with minimal regeneration as a conse- figures are copyrighted by and modified with permission quence. The third photomicrograph shows a sham-sur- from the American Academy of Periodontology.

Fig. 9. Critical-size, 5-mm, supraalveolar periodontal de- a functionally oriented periodontal ligament, and alveolar fect implanted with an occlusive, space-providing expan- bone. Note the gradual thinning of the regenerated cel- ded polytetrafluoroethylene (ePTFE) membrane. The lular cementum in a coronal direction. Healing interval high-magnification photomicrographs from the apical, 8 weeks. For detail see: Sigurdsson et al. 1994 (38). These mid, and coronal aspect of the defect show regeneration of figures are copyrighted by and modified with permission the periodontal attachment, including cellular cementum, from the American Academy of Periodontology. however, remarkable clinical differences between the including primary intention healing and space pro- experimental conditions. Whereas all sites receiving vision without barrier membranes. the porous membrane remained submerged for primary intention healing, 50% of sites receiving the Primary intention healing vs. wound occlusive membrane exhibited wound failure and failure and membrane exposure membrane exposure. Obviously the porous membrane supported flap survival, probably being less of Wound failure including membrane exposure is a a challenge to the vascular support of the gingival calamity of periodontal-regenerative therapy utilizing flaps than the occlusive membrane. The results of GTR techniques, making the procedure unpredicta- this study ultimately support a concept of periodon- ble in clinical practice (35, 41). The membrane can be tal regeneration following gingival flap surgery, difficult to submerge completely by gingival tissues at

38 Biology and principles of periodontal wound healing/regeneration

Fig. 10. Critical-size, 5-mm, supraalveolar periodontal Similar results were found in sites implanted with the defect implanted with occlusive and macro-porous, space- occlusive membrane, clearly suggesting that tissue providing expanded polytetrafluoroethylene (ePTFE) occlusion is not a critical requirement for periodontal membranes; the photomicrographs show a site implanted regeneration. Healing interval 8 weeks. For detail see: with the porous membrane. Note significant periodontal Wikesjo¨ et al. 2003 (52). These figures are copyrighted by regeneration, including a functionally oriented perio- and modified with permission from Blackwell Munks- dontal ligament, cellular cementum, and alveolar bone, gaard. approaching the cemento–enamel junction (green arrow). wound closure, or it may exhibit subclinical exposure have been demonstrated in a retrospective evaluation or poor flap retention, even following the best of GTR therapy in 38 healthy patients receiving intentions for primary intention healing, and thus treatment of intrabony periodontal defects with a becomes exposed during the healing sequel. Clinical defect depth averaging 6.5 ± 1.6 mm and probing experience and histologic evaluations of periodontal depth averaging 7.6 ± 1.5 mm (41). Probing bone wound healing in supraalveolar periodontal defects level gain in sites without membrane exposure aver- demonstrate that GTR membranes frequently aged 4.1 ± 2.3 mm, in contrast to 2.2 ± 2.3 mm for become exposed, possibly as a consequence of sites with membrane exposure. These observations compromised nutritional support to the overlaying likely apply to all membrane technologies until gingival tissues (38, 52). Oral bacteria, provoking an shown otherwise. The observations demonstrate the inflammatory reaction within the regenerate under- critical significance of primary (unexposed) intention neath the membrane, in turn colonize the exposed healing for periodontal regeneration. sites. In early studies, animals experiencing membrane exposure received systemic antimicrobial Defect characteristics, space provision, therapy and daily rinses with a chlorhexidine gluco- and innate regenerative potential nate solution throughout the healing sequel. Al- though this treatment maintains optimal gingival Defect configuration is considered to be a critical health in nonexposed sites, exposed sites exhibit factor influencing the outcome of periodontal- large inflammatory infiltrates and necrotic tissues regenerative therapy in clinical practice. Deep, with limited, if any, signs of periodontal regeneration narrow intrabony defects appear to be favorable in the histological evaluation (Fig 7 and 8). In con- candidates for regenerative surgery compared with trast, when GTR membranes were removed imme- wide, shallow defects (6), as do three-wall intrabony diately upon exposure followed by wound closure defects compared with two- and one-wall intrabony over the exposed regenerate, the newly formed tis- defects. Supracrestal periodontal regeneration is sues matured into alveolar bone, cementum, and a generally not considered a clinical possibility. From functionally oriented periodontal ligament, even in a conceptual point of view, it appears logical that sites where the wound failure/membrane exposure deep, narrow, three-wall intrabony defects should occurred as early as 1 week postsurgery (52). In sites react favorably over shallower, wider, and more where periodontal regeneration is allowed to pro- open sites. The relative abundance of tissue re- gress unobstructed under conditions for primary sources contributing to the regeneration in three- intention healing, complete, or almost complete, wall intrabony defects, the defect area being more or regeneration of the periodontal attachment becomes less circumscribed by the residual periodontal liga- an imminent possibility (Fig 8–10) (38, 51–53). The ment, should enhance the regenerative potential of clinical significance of these biologic observations these sites over that in two- and one-wall intrabony

39 Polimeni et al. defects, providing that conditions for primary intention healing are maintained. However, observations of periodontal wound healing and regeneration in this text, based on the Critical-size Supraalveolar Periodontal Defect Model, demonstrate the biologic possibility of extensive, if not complete, regeneration of the periodontal attachment, including alveolar bone, in supracrestal, Ôzero-wallÕ periodontal defects (Fig 8–10) (38, 51–53). Early reports, evaluating GTR technology using barrier membranes and supraalveolar periodontal defects, point to a key role of space provision in Fig. 11. Photomicrograph showing parameters evaluated periodontal-regenerative therapy. Haney et al. (9) by Polimeni et al., including defect area (orange) underneath the membrane delineated by the base of the defect reported a significant correlation (r ¼ 0.997; P ¼ apically and the cemento–enamel junction coronally, 0.002) between the space provided by the membrane bone regeneration height (green arrow), the height of and the newly formed bone (Fig. 7). Sigurdsson et al. the newly formed alveolar bone along the root surface (38) showed that sites subject to space provision representing a surrogate parameter for regeneration exhibited extensive bone regeneration compared of the periodontal attachment (G. Polimeni & C. Susin unpublished); and the width of the alveolar crest (yellow with that in controls (Fig. 8). In other words, a large arrow) at the base of the defect. For detail see: Polimeni wound area resulted in increased bone regeneration. et al. 2004 (32); Polimeni et al. 2004 (29); Polimeni et al. Sigurdsson et al. (38) reported greater bone regener- 2004 (30); Polimeni et al. 2004 (31); Polimeni et al. 2005 ation compared with that reported by Haney et al. (9), (33). This figure is copyrighted by and published with although the space underneath the GTR membrane permission from Blackwell Munksgaard. was not completely filled with alveolar bone. Rather, the newly formed bone assumed a ÔphysiologicÕ form The effect of defect characteristics and space pro- along the root surface, much like the preceding vision, and innate regenerative potential has received resident bone, and the remaining space underneath further analysis using the Critical-size Supraalveolar the membrane was occupied by dense connective Periodontal Defect Model. Polimeni et al. (29–33) used tissue (Figs 8 and 9). There were, however, meth- the height of the regenerated alveolar bone along the odological variations between the studies that may root surface as a parameter for periodontal regener- account for the differences observed in the magni- ation to evaluate the biologic potential for regener- tude of alveolar bone regeneration. Haney et al. (9) ation under various conditions. Other parameters used transgingival wound closure and positioned included the width of the alveolar crest at the base of individual clinical GTR membranes around the neck the defect and the wound area delineated by the base of the teeth, while Sigurdsson et al. (38) positioned a of the defect, the lateral extension of a GTR mem- structurally reinforced space-providing membrane in brane, the cemento–enamel junction, and the tooth such a manner that the teeth and membrane became surface (Fig. 11). The use of the height of the regen- completely submerged, suturing the flaps over the erated alveolar bone as a surrogate parameter for top of the membrane. The wound area delineated by periodontal regeneration was based on observations the membrane was obviously increased in the latter suggesting a significant correlation between the study. It can be speculated that variation in posi- height of newly formed bone along the root surface tioning of the membrane influenced space provision and regeneration of the periodontal attachment and that this, in turn, influenced the regenerative extending just coronally of the alveolar crest in supra- potential of the defect sites. Similar observations alveolar periodontal defects (r ¼ 0.96; P ¼ 0.001; G. have been reported by Cortellini et al. (7) and Tonetti Polimeni & C. Susin 2005, unpublished). These et al. (40) in clinical studies aimed at investigating observations as such suggest that the periodontal factors affecting the healing response of intrabony ligament provides the leading edge for the regene- defects following GTR and access flap surgery. Space rating periodontal tissues. provision and wound stability have also been A first study evaluated the role of space-provision advocated as main factors influencing the magnitude for periodontal regeneration. Supraalveolar, period- of bone regeneration in three-wall compared with ontal defects, having received either a space-provid- two- and one-wall intrabony defects in a dog model ing, porous ePTFE membrane or sham-surgery, were (16). subject to histometric analyses (including investiga-

40 Biology and principles of periodontal wound healing/regeneration tion of vertical regeneration of the alveolar bone and GTR procedures, including the placement of a analysis of the width of the alveolar crest at the base nonresorbable bone biomaterial in supraalveolar of the defect) following an 8-week healing interval periodontal defects (Fig. 12). They found a signifi- (33). Bone regeneration at sites receiving the space- cantly positive correlation between space provision providing membrane was significantly greater than by the membrane and alveolar regeneration, and a that at sites receiving sham-surgery (P ¼ 0.0003). A significantly negative correlation between the density significant relationship between the width of the of biomaterials and alveolar regeneration (P < 0.01). alveolar crest at the base of the defect and bone In other words, the biomaterial obstructed the space, regeneration was observed, with no significant dif- thus preventing regeneration. Polimeni et al (32). ference between sites receiving different treatments evaluated periodontal regeneration following GTR (P ¼ 0.84). It can be concluded therefore that space procedures, including placement of a resorbable provision has a significant effect on periodontal bone biomaterial (Fig. 13). Bilateral, supraalveolar, regeneration. Notably, the width of the alveolar crest periodontal defects, having received a clinical ePTFE at the base of the defect appears to influence space membrane with (cGTR) or without (GTR) the coral- provision effectively, supporting regeneration. Sites derived biomaterial, were subject to histometric providing a wide alveolar base showed enhanced analysis, including vertical regeneration of alveolar regeneration, whereas sites exhibiting a narrow base bone relative to space provision by the ePTFE showed limited regeneration for both treatment membrane, following a 4-week healing interval. Sig- conditions. One may speculate that in the presence of nificantly greater bone regeneration was observed at a wide alveolar base, the mucoperiosteal flap serves sites receiving cGTR compared with GTR alone the same mechanical function as the space-provi- (P < 0.0001). Sites providing larger wound areas ding, porous ePTFE membrane, whereas in the showed greater bone regeneration compared with presence of a narrow base, the flap and the mem- sites exhibiting smaller wound areas (P < 0.0001). brane-supported flap collapse onto the tooth surface, However, grouping the sites by wound area thres- providing limited space for regeneration. In other holds showed that bone regeneration was not sig- words, the characteristics of the mucoperiosteal flap nificantly different in sites receiving cGTR compared alone, or supported by the space-providing, porous with GTR alone, irrespective of the size of the wound ePTFE membrane, are not different, from a wound area (P > 0.5). This study showed that the coral-de- mechanical point of view. rived biomaterial influences space provision by en- A clinical approach to space provision for perio- hancing the wound area. The physical structure of dontal regeneration has included the placement of the biomaterial appeared to prevent the GTR mem- bone biomaterials to support GTR membranes. brane from collapsing onto the root surface. This Trombelli et al. (42) evaluated regeneration following overall effect supported enhanced bone formation in

Fig. 12. Critical-size, 5-mm, supraalveolar periodontal bone biomaterials in conjunction with periodontal- defect implanted with an occlusive expanded polytetra- regenerative surgery. This study points to the critical fluoroethylene (ePTFE) membrane and an osteoconduc- importance of unobstructed space provision for perio- tive biomaterial. Note significant regeneration of alveolar dontal regeneration. Healing interval 4 weeks. For detail bone in the left photomicrograph in the absence of the see: Trombelli et al. 1999 (42). These figures are copy- biomaterial. It is shown that less tissue regeneration righted by and modified with permission from Blackwell occurs as more biomaterial is used, raising the question of Munksgaard. any benefit of using slowly resorbing or nonresorbable

41 Polimeni et al.

Fig. 13. Critical-size, 5-mm, supraalveolar periodontal sites, bone regeneration appears to be influenced by space defects implanted with an occlusive expanded polytetra- provision in the GTR sites. This study showed that the fluoroethylene (ePTFE) membrane for guided tissue coral-derived biomaterial influences space provision by regeneration (GTR) in the presence (cGTR) or absence enhancing the wound area. The physical structure of the (GTR) of a coral-derived biomaterial. The green arrow- biomaterial appeared to prevent the GTR membrane from heads delineate the apical aspect of the supraalveolar collapsing onto the root surface. This overall effect sup- periodontal defect. The left photomicrograph shows a site ported enhanced bone formation in sites receiving cGTR receiving cGTR where the membrane has been com- compared with sites receiving GTR alone. Importantly, pressed onto the root, with minimal regeneration as a when adjusted for the effect of wound area, a two-way consequence. The left center photomicrograph shows a analysis of variance (ANOVA) did not show statistically site receiving the same treatment protocol with enhanced significant differences between the protocols. Healing space provision allowing increased bone regeneration. interval 4 weeks. For detail see: Polimeni et al. 2004 (32). The right and right center photomicrographs show sites These figures are copyrighted by and modified with per- receiving GTR alone. Similarly to that observed for cGTR mission from Blackwell Munksgaard.

sites receiving cGTR compared with sites receiving to provide for GTR in supraalveolar periodontal de- GTR alone. However, when adjusted for the effect of fects (Fig. 14). The gingival flaps were advanced for wound area, a two-way analysis of variance did not primary intention healing that was allowed to pro- show statistically significant differences between the gress for 8 weeks. The histometric analysis assessed protocols. Consistent with this observation, stratifi- regeneration of alveolar bone relative to space pro- cation of the wound area into subgroups did not re- vision by the ePTFE membranes. The bivariate ana- veal significant differences between the protocols. lysis showed that space provision and membrane This should be interpreted to indicate that the coral- occlusivity significantly enhanced bone regeneration. derived biomaterial did not exhibit osteoconductive Sites receiving the occlusive GTR membrane, and properties. In other words, it did not enhance the sites with enhanced space provision, showed signifi- osteogenic potential of the site. These observations cantly greater bone regeneration than sites receiving corroborate histopathological evaluations of this the porous GTR membrane (P ¼ 0.03) or exhibiting biomaterial in periodontal sites (19). On the other more limited space provision (P ¼ 0.0002). Never- hand, the coral-derived biomaterial did not appear to theless, a significant relationship was observed be- obstruct regeneration, in contrast to that observed for tween bone regeneration and space provision for other particulate biomaterials used to support space sites receiving the occlusive (b ¼ 0.194; P < 0.02) and provision or to serve as osteoconductive conduits the porous (b ¼ 0.229; P < 0.0004) GTR membranes, in conjunction with GTR (39, 42). Future research, irrespective of treatment (P ¼ 0.14). In other words, evaluating osteoconductive properties of biomateri- the relationship between space provision and re- als to be used for space provision/regeneration, generation was significant for both the porous and should take into consideration the native osteogenic the occlusive GTR membranes. Regeneration fol- potential. Proper methodology and analysis should lowed similar patterns in both groups. It may be be applied to distinguish this effect from any osteo- speculated that the healing process supported by conductive effects of the biomaterial. these different membranes is similar, or at least In a separate evaluation, Polimeni et al. (29) esti- similarly influenced by space provision. Nevertheless, mated the effect of cell occlusion and space provision the magnitude of regeneration was significantly in- on periodontal regeneration. Space-providing occlu- creased at sites receiving the occlusive GTR mem- sive and porous ePTFE membranes were implanted branes compared with that at sites receiving the

42 Biology and principles of periodontal wound healing/regeneration

Fig. 14. Critical-size, 5-mm, supraalveolar periodontal eration compared with sites receiving the porous GTR defect implanted with an occlusive, space-providing membrane (P ¼ 0.03) or exhibiting more limited space expanded polytetrafluoroethylene (ePTFE) membrane (A) provision (P ¼ 0.0002). Nevertheless, the relationship be- and with a porous ePTFE membrane (B). Green arrow- tween space provision and regeneration was significant heads delineate the apical aspect of the supraalveolar for both occlusive and porous GTR membranes. Regen- periodontal defects. Green lines approximate the coronal eration followed similar patterns for both groups. It may aspect of the regenerated bone. Notably, bone regener- be speculated that the healing process supported by these ation is influenced by space provision under the mem- different membranes may be similar to, or at least be branes. This study showed that space provision and similarly influenced by, space provision. Healing interval membrane occlusivity significantly enhanced bone 8 weeks. For detail see: Polimeni et al. 2004 (29). These regeneration. Sites receiving the occlusive guided tissue figures are copyrighted by and modified with permission regeneration (GTR) membrane and sites with enhanced from Blackwell Munksgaard. space provision showed significantly greater bone regen-

porous GTR membrane, when adjusted for the effect after which block sections were collected for histo- of wound area. Thus, even if space provision appears metric analysis, including analysis of regeneration of to be a critical factor for regeneration, membrane alveolar bone relative to space provision by the GTR occlusivity appears to provide adjunctive effects. membrane and width of the alveolar crest at the base While it may not be legitimate to consider cell oc- of the defect. There were no significant differences in clusion as an absolute prerequisite for periodontal mean alveolar regeneration between sites receiving regeneration (52), it appears that the use of cell- the porous GTR membrane with a narrow vs. a wide occlusive membranes may optimize the magnitude alveolar base after adjusting for wound area (2.2 vs. of regeneration. 2.5 mm, respectively; P ¼ 0.36). In contrast, analysis The influence of the resident alveolar bone on bone using sites receiving the occlusive GTR membrane regeneration in conjunction with GTR, in the pres- revealed significantly greater bone regeneration at ence or absence of cell occlusivity, was evaluated in a sites with a wide compared with a narrow alveolar subsequent analysis (31). Space-providing, occlusive base (3.3 vs. 2.5 mm, respectively; P ¼ 0.02). Regres- or porous ePTFE membranes were implanted into sion analysis showed a significant relationship contralateral supraalveolar periodontal defects to (P ¼ 0.05) between space provision and bone regen- assist GTR under conditions for primary intention eration for all groups, except for sites with a wide healing (Fig. 14). The healing interval was 8 weeks, alveolar base receiving the occlusive GTR membrane

43 Polimeni et al.

Fig. 15. Critical-size, 5-mm, supraalveolar peri-implant cantly influences alveolar bone regeneration. Adjusting defect including three titanium implants and a space- for the effect of wound area, periodontal sites exhibit providing porous expanded polytetrafluoroethylene significantly increased bone regeneration compared with (ePTFE) membrane. The green arrowheads delineate the that in alveolar (peri-implant) sites. This observation apical aspect of the supraalveolar peri-implant defect. The suggests critical biologic differences between periodontal red line approximates the coronal aspect of the newly and alveolar (peri-implant) sites. Healing interval 8 weeks. formed bone. Notably, the amount of bone regeneration is For detail see: Polimeni et al. 2004 (30). These figures are influenced by space provision under the membrane. This copyrighted by and modified with permission from study showed that the width of the alveolar crest and the Blackwell Munksgaard. space provided by the porous ePTFE membrane signifi-

(P ¼ 0.5). The present study (undertaken in the experimental sites had been subject to GTR using presence of tissue occlusion and controlling for space-providing porous ePTFE membranes. The wound area) established that sites exhibiting a healing interval was 8 weeks. The histometric analy- wide alveolar base might have a greater osteogenic sis assessed alveolar bone regeneration (height) rel- potential than sites with a narrow base. This obser- ative to space provision by the membrane and the vation suggests that the osteogenic potential of the width of the alveolar crest at the base of the defect. resident bone plays a role in periodontal regener- Statistical analysis used the linear mixed models. The ation. On the other hand, in the absence of tissue results revealed a significant correlation between occlusion (porous GTR membranes) and controlling bone width and wound area (r ¼ 0.56, P < 0.0001). for wound area, sites exhibiting a wide alveolar base Generally, bone width and wound area had statisti- did not show an enhanced osteogenic potential cally significant effects on the extent of bone re- compared to sites with a narrow base. One may generation (P < 0.0005 and P < 0.0001, respectively). speculate that this might be a consequence of the Bone regeneration was linearly correlated with the porous space-providing GTR membrane allowing the bone width at periodontal (P < 0.001) and implant gingival connective tissue access to the wound area. (P ¼ 0.04) sites, and with the wound area at period- Consequently, tissue resources, including molecules, ontal (P < 0.0001) and implant (P ¼ 0.03) sites. The cells, and vascularity originating from the gingival relationships of bone regeneration with these two connective tissue, may have an inhibitory effect on variables were not significantly different between osteogenesis, and/or migration and proliferation of teeth and implants (bone width, P ¼ 0.83; wound tissue elements from the gingival connective tissue area, P ¼ 0.09). When adjusted for wound area, bone competitively occupied the space for bone to form regeneration was significantly greater at periodontal into. Thus, the resident alveolar bone might signifi- than at implant sites (P ¼ 0.047). Thus, the histo- cantly influence the magnitude of alveolar bone metric analysis suggested similar patterns of bone regeneration, while the relative presence of cells from regeneration at periodontal and implant sites. Simi- the gingival connective tissue may attenuate this larities in the behavior of factors influencing bone effect. regeneration were observed for both periodontal and Subsequently, Polimeni et al. (30) evaluated the implant sites. The width of the alveolar crest and the influence of alveolar bone and space provision on space provided by the porous ePTFE membrane bone regeneration at teeth and titanium implants, resulted in a significant relationship with the extent comparing observations at supraalveolar periodontal of alveolar bone regeneration for both sites. Adjusting and supraalveolar peri-implant defects (Fig. 15). The for the effect of wound area, the periodontal sites

44 Biology and principles of periodontal wound healing/regeneration exhibited significantly greater bone regeneration than References the implant sites. This observation suggests critical biologic differences between periodontal and implant 1. Aukhil I. Biology of wound healing. Periodontol 2000 2000: sites. While bone regeneration in periodontal sites 22: 44–50. may be induced by vascular and cellular elements 2. Baker PJ, Rotch HA, Trombelli L, Wikesjo¨ UME. An in vitro sequestered in the periodontal ligament, or by screening model to evaluate root conditioning protocols for periodontal regenerative procedures. J Periodontol 2000: synergistic effects between the periodontal attach- 71: 1139–1143. ment and the resident alveolar bone, regeneration at 3. Baker DL, Stanley Pavlow SA, Wikesjo¨ UME. Fibrin clot implant sites appears to be solely dependent on the adhesion to dentin conditioned with protein constructs: an evidently limited regenerative potential of the alveo- in vitro proof-of-principle study. J Clin Periodontol 2005: lar bone. In consequence, principles valid for GTR in 32: 561–566. 4. Caton J, Nyman S, Zander H. Histometric evaluation of periodontal sites may not necessarily immediately periodontal surgery. II. Connective tissue attachment levels apply to implants. Moreover, from a clinical per- after four regenerative procedures. J Clin Periodontol 1980: spective, these biologic observations suggest that 7: 224–231. bone-regenerative procedures at implant sites may 5. Clark RAF. Wound repair. Overview and general consider- be considerably more challenging than at periodontal ations. In: Clark RAF, editor. The Molecular and Cellular Biology of Wound Repair, 2nd edn. New York, NY: Plenum sites. Press, 1996: 3–50. 6. Cortellini P, Tonetti M. Focus on intrabony defects: guided tissue regeneration. Periodontol 2000 2000: 22: 104– 132. Conclusions 7. Cortellini P, Pini Prato G, Tonetti MS. Periodontal regeneration of human intrabony defects with titanium rein- Current scientific evidence points to the presence of forced membranes. A controlled clinical trial. J Periodontol cells originating from the periodontal ligament, 1995: 66: 797–803. wound stability, space provision, and primary inten- 8. Grzesik WJ, Narayanan AS. Cementum and periodontal tion healing, as fundamental biologic and clinical wound healing and regeneration. Crit Rev Oral Biol Med 2002: 13: 474–484. factors that must be met to obtain periodontal 9. Haney JM, Nilveús RE, McMillan PJ, Wikesjo¨ UME. Perio- regeneration. dontal repair in dogs: expanded polytetrafluoroethylene Only a profound understanding of biological and barrier membranes support wound stabilization and en- clinical variables affecting the outcome of perio- hance bone regeneration. J Periodontol 1993: 64: 883–890. dontal-regenerative procedures will allow clinicians 10. Hardwick R, Hayes BK, Flynn C. Devices for dentoalveolar regeneration: an up-to-date literature review. J Periodontol to manipulate biological and clinical factors effect- 1995: 66: 495–505. ively in order to optimize the clinical result and in- 11. Hiatt WH, Stallard RE, Butler ED, Badgett B. Repair fol- crease the predictability of periodontal-regenerative lowing mucoperiosteal flap surgery with full gingival therapy. retention. J Periodontol 1968: 39: 11–16. It is evident that the search for novel regenerative 12. Karaki R, Kubota K, Hitaka M, Yamaji S, Kataoka R, Yamamoto H. Effect of gum-expanding mesh on the therapies requires preclinical evaluation in well- osteogenesis in surgical bony defect. Nippon Shishubyo characterized rodent screening models to determine Gakkai Kaishi 1984: 26: 516–522. biologic potential and safety, and analysis in 13. Karring T, Nyman S, Lindhe J. Healing following implan- discriminating large animal models (designated as tation of periodontitis affected roots into bone tissue. J Clin Critical-size Defect Models) to establish clinical Periodontol 1980: 7: 96–105. 14. Karring T, Nyman S, Gottlow J, Laurell L. Development of potential and efficacy. Clinical evaluation should be the biological concept of guided tissue regeneration–Ani- limited to therapies showing promising results in mal and human studies. Periodontol 2000: 1993: 26–35. these preclinical models. 15. Kim CK, Cho KS, Choi SH, Prewett A, Wikesjo¨ UME. Peri- odontal repair in dogs: effect of allogenic freeze-dried demineralized bone matrix implants on alveolar bone and cementum regeneration. J Periodontol 1998: 69: 26–33. Acknowledgment 16. Kim CS, Choi SH, Chai JK, Cho KS, Moon IS, Wikesjo¨ UME, Kim CK. Periodontal repair in surgically created intrabony Earlier versions of this text have been published defects in dogs: Influence of the number of bone walls on for reviews in journals and book chapters. The text healing response. J Periodontol 2004: 75: 229–235. 17. Koo K-T, Polimeni G, Albandar JM, Wikesjo¨ UME. Perio- is continuously subject to revisions and updating dontal repair in dogs: examiner reproducibility in the as new information becomes available in our supraalveolar periodontal defect model. J Clin Periodontol laboratory. 2004: 31: 439–442.

45 Polimeni et al.

18. Koo K-T, Polimeni G, Albandar JM, Wikesjo¨ UME. Perio- membranes: results from a multicenter practice-based dontal repair in dogs: analysis of histometric assessments clinical trial. J Periodontol 2004: 75: 726–733. in the supraalveolar periodontal defect model. J Periodon- 36. Scantlebury TV. 1982–1992: a decade of technology devel- tol 2004: 75: 1688–1693. opment for guided tissue regeneration. J Periodontol 1993: 19. Koo K-T, Polimeni G, Qahash M, Kim CK, Wikesjo¨ UME. 64: 1129–1137. Periodontal repair in dogs: guided tissue regeneration 37. Selvig KA, Biagiotti GR, Leknes KN, Wikesjo¨ UME. Oral enhances bone formation in sites implanted with a coral- tissue reactions to suture materials. Int J Periodontics derived calcium carbonate biomaterial. J Clin Periodontol Restorative Dent 1998: 18: 474–487. 2005: 32: 104–110. 38. Sigurdsson TJ, Hardwick R, Bogle GC, Wikesjo¨ UME. 20. Kornman KS, Robertson PB. Fundamental principles Periodontal repair in dogs: space provision by reinforced affecting the outcomes of therapy for osseous lesions. ePTFE membranes enhances bone and cementum regen- Periodontol 2000 2000: 22: 22–43. eration in large supraalveolar defects. J Periodontol 1994: 21. Leknes KN, Selvig KA, Bøe OE, Wikesjo¨ UME, Tissue 65: 350–356. reactions to sutures in the presence and absence of 39. Stavropoulos A, Kostopoulos L, Mardas N, Nyengaard JR, anti-infective therapy. J Clin Periodontol 2005: 32: 130– Karring T. Deproteinized bovine bone used as an 138. adjunct to guided bone augmentation: an experimental 22. Linghorne WJ, O’Connell DC. Studies in the regeneration study in the rat. Clin Implant Dent Relat Res 2001: 3: and reattachment of supporting structures of teeth. I. Soft 156–165. tissue reattachment. J Dent Res 1950: 29: 419–428. 40. Tonetti MS, Pini Prato G, Cortellini P. Factors affecting the 23. Listgarten MA, Rosenberg MM. Histological study of repair healing response of intrabony defects following guided following new attachment procedures in human perio- tissue regeneration and access ﬂap surgery. J Clin Period- dontal lesions. J Periodontol 1979: 50: 333–344. ontol 1996: 23: 548–556. 24. Martin P. Wound healing – aiming for perfect skin regen- 41. Trombelli L, Kim CK, Zimmerman GJ, Wikesjo¨ UME. Ret- eration. Science 1997: 276: 75–81. rospective analysis of factors related to clinical outcome of 25. Melcher AH. On the repair potential of periodontal tissues. guided tissue regeneration procedures in intrabony defects. J Periodontol 1976: 47: 256–260. J Clin Periodontol 1997; 24: 366–371. 26. Nyman S, Karring T, Lindhe J, Planten S. Healing fol- 42. Trombelli L, Lee MB, Promsudthi A, Guglielmoni PG, Wi- lowing implantation of periodontitis-affected roots into kesjo¨ UME. Periodontal repair in dogs: histologic observa-

gingival connective tissue. J Clin Periodontol 1980: 7: 394– tions of guided tissue regeneration with a prostaglandin E1 401. analog/methacrylate composite. J Clin Periodontol 1999: 27. Nyman S, Gottlow J, Karring T, Lindhe J. The regener- 26: 381–387. ative potential of the periodontal ligament. An experi- 43. Wikesjo¨ UME, Nilveús R. Periodontal repair in dogs: effect mental study in the monkey. J Clin Periodontol 1982: 9: of wound stabilization on healing. J Periodontol 1990: 61: 257–265. 719–724. 28. Nyman S, Lindhe J, Karring T, Rylander H. New attachment 44. Wikesjo¨ UME, Nilveús R. Periodontal repair in dogs: heal- following surgical treatment of human periodontal disease. ing patterns in large circumferential periodontal defects. J Clin Periodontol 1982: 9: 290–296. J Clin Periodontol 1991: 18: 49–59. 29. Polimeni G, Koo K-T, Qahash M, Xiropaidis AV, Albandar 45. Wikesjo¨ UME, Selvig KA. Periodontal wound healing and JM, Wikesjo¨ UME. Prognostic factors for alveolar regener- regeneration. Periodontol 2000 1999: 19: 21–39. ation: effect of tissue occlusion on alveolar bone regener- 46. Wikesjo¨ UME, Claffey N, Christersson LA, Franzetti LC, ation with guided tissue regeneration. J Clin Periodontol Genco RJ, Terranova VP, Egelberg J. Repair of periodontal 2004: 31: 730–735. furcation defects in beagle dogs following reconstructive 30. Polimeni G, Koo K-T, Qahash M, Xiropaidis AV, Albandar surgery including root surface demineralization with tet- JM, Wikesjo¨ UME. Prognostic factors for alveolar regener- racycline hydrochloride and topical fibronectin applica- ation: bone formation at teeth and titanium implants. J Clin tion. J Clin Periodontol 1988: 15: 73–80. Periodontol 2004: 31: 927–932. 47. Wikesjo¨ UME, Claffey N, Nilveús R, Egelberg J. Periodontal 31. Polimeni G, Albandar JM, Wikesjo¨ UME. Prognostic factors repair in dogs: effect of root surface treatment with stan- for alveolar regeneration: osteogenic potential of resident nous fluoride or citric acid on root resorption. J Periodontol bone. J Clin Periodontol 2004: 31: 840–844. 1991: 62: 180–184. 32. Polimeni G, Koo K-T, Qahash M, Xiropaidis AV, Albandar 48. Wikesjo¨ UME, Claffey N, Egelberg J. Periodontal repair in JM, Wikesjo¨ UME. Prognostic factors for alveolar regen- dogs: effect of heparin treatment of the root surface. J Clin eration: effect of a space-providing biomaterial on guided Periodontol 1991: 18: 60–64. tissue regeneration. J Clin Periodontol 2004: 31: 725– 49. Wikesjo¨ UME, Crigger M, Nilveús R, Selvig KA. Early healing 729. events at the dentin-connective tissue interface. Light and 33. Polimeni G, Albandar JM, Wikesjo¨ UME. Prognostic factors transmission electron microscopy observations. J Period- for alveolar regeneration: effect of space provision. J Clin ontol 1991: 62: 5–14. Periodontol 2005: 32: 951–954. 50. Wikesjo¨ UME, Kean CJC, Zimmerman GJ. Periodontal re- 34. Polson AM, Proye MP. Fibrin linkage: a precursor for new pair in dogs: Supraalveolar defect models for evaluation of attachment. J Periodontol 1983: 54: 141–147. safety and efficacy of periodontal reconstructive therapy. 35. Sanz M, Tonetti MS, Zabalegui I, Sicilia A, Blanco J, Rebelo J Periodontol 1994: 65: 1151–1157. H, Rasperini G, Merli M, Cortellini P, Suvan JE. Treatment 51. Wikesjo¨ UME, Xiropaidis AV, Thomson RC, Cook AD, Selvig of intrabony defects with enamel matrix proteins or barrier KA, Hardwick WR. Periodontal repair in dogs: rhBMP-2

46 Biology and principles of periodontal wound healing/regeneration

signiﬁcantly enhances bone formation under provisions for a bioresorbable space-providing macro-porous membrane guided tissue regeneration. J Clin Periodontol 2003: 30: with recombinant human bone morphogenetic protein-2. 705–714. J Periodontol 2003: 74: 635–647. 52. Wikesjo¨ UME, Lim WH, Thomson RC, Hardwick WR. 54. Wikesjo¨ UME, Lim WH, Razi SS, Sigurdsson TJ, Lee MB, Periodontal repair in dogs: gingival tissue occlusion, a Tatakis DN, Hardwick WR. Periodontal repair in dogs: a critical requirement for guided tissue regeneration? J Clin bioresorbable calcium carbonate coral implant enhances Periodontol 2003: 30: 655–664. space provision for alveolar bone regeneration in con- 53. Wikesjo¨ UME, Lim WH, Thomson RC, Cook AD, Wozney junction with guided tissue regeneration. J Periodontol JM, Hardwick WR. Periodontal repair in dogs: evaluation of 2003: 74: 955–962.