Articles
Søren JEPSEN * Hendrick TERHEYDEN **
*Department of Periodontics, Kiel University, Germany
**Department of Oral-Maxillofacial Surgery, Kiel University, Germany
Polypeptide growth and differentiation factors (GFs) are naturally occurring biologic mediators that regulate key events in tissue repair such as chemotaxis, differentiation, proliferation and matrix synthesis of cells. They act mainly in an autocrine (where cells that produce GFs are also affected by the same GFs) or paracrine fashion (the production of GFs by one cell type affects the function of a different cell type) by binding to specific cell surface receptors...
Polypeptide growth factors and morphogenetic proteins are biological mediators which play an important role in the stimulation and regulation of wound healing. These factors are involved in repair and regeneration by regulating mitogenesis, differentiation and matrix synthesis. Numerous preclinical studies have demonstrated the potential of certain growth and differentiation factors for the modulation of periodontal and periimplant wound healing resulting in substantial periodontal regeneration and enhanced oral implant integration. Several peptides are available in recombinant form and could be utilized in therapeutic settings. It is anticipated, that in the future the application of these very potent growth proteins will become part of our regenerative strategies. This review will focus on our current understanding of the effects of growth factors and morphogenetic proteins on cells and tissues involved in periodontal regeneration and in the osseous integration of oral implants.
Polypeptide growth and differentiation factors (GFs) are naturally occurring biologic mediators that regulate key events in tissue repair such as chemotaxis, differentiation, proliferation and matrix synthesis of cells. They act mainly in an autocrine (where cells that produce GFs are also affected by the same GFs) or paracrine fashion (the production of GFs by one cell type affects the function of a different cell type) by binding to specific cell surface receptors using signal transducting pathways that involve tyrosine kinase or serine/thyronase phosphorylation. Examples for growth and differentiation factors found in bone and healing tissues include PDGF, TGF-β, FGF, IGF-I and -II and bone morphogenetic proteins (BMPs). In this context, growth factors are polypeptides that have an effect on the growth of cells (mitogenic polypeptides) and in addtition they may have an effect on the rate of extracellular matrix production. In contrast, differentiation factors (differentiation polypeptides = morphogenetic proteins) control the phenotype of cells, causing stem cells or progenitor cells to become committed cells (i.e. the differentiation of undifferentiated mesenchymal cells to osteoblasts) (fig. 1).
Wound healing is thought to be regulated by various growth and differentiation factors and cytokines. Following acute tissue injury activated platelets at the wound margins release growth factors such as PDGF and TGF-β1 (Assoian et al., 1984). The plasma exudate is an important source of insulin-like growth factors (Rinderknecht and Humbel, 1978). Cells adjacent to the site of injury release growth factors such as IGF-1, PDGF and TGF-β1 within a few hours after injury (Hansson et al., 1987 ; Sitaras et al., 1987 ; Antoniades et al., 1991 ; Werner et al., 1992). Macrophages entering the wound healing arena serve as another source for PDGF and TGF-β1 (Rappolee et al., 1988). Bone is known to be a major storage site of growth factors such as IGF-I, -II, TGF-β and PDGF which may also be released during injury (Hauschka, 1990). Bone morphogenetic proteins (BMPs) are also present in bone matrix (Urist, 1965, Urist and Strates, 1971) and expressed during fracture repair (Nakase et al., 1994). The expression of GFs following bone and soft tisue injury is thought to regulate the process of repair and/or regeneration. Some authors have even hypothesized that periodontal breakdown could be a consequence of inadequate repair (Hausmann et al., 1992). Mediators of inflammation may have an inhibitory effect on growth factors such as PDGF (McAllister et al., 1995). Thus, the rationale for growth factor administration in periodontics and oral implant dentistry is to enhance and/or accelerate the physiological wound healing capacity, that may be insufficient to promote a complete healing of the affected structures. Over the past decade numerous in vitro and in vivo studies have been performed to elucidate the role of growth and differentiation factors such as PDGF and BMP on periodontal wound healing and oral implant integration. Several of these factors are now available in recombinant form and can be produced in a highly pure form in a large scale production (fig. 2a, 2b, 2c and 2d).
This review will describe the effects of growth factors and morphogenetic proteins on periodontal ligament fibroblasts and hard tissue structures cementum and alveolar bone based on available in vitro and in vivo studies.
Factors that possess stimulating effects on the proliferation, migration and collagen matrix synthesis of periodontal ligament (PDL) fibroblasts cells may have the potential to promote new attachment formation. At present there is information available regarding effects on PDL cell activity for PDGF, TGF-β1, IGF-I, CGF and BMP-7.
Platelet-Derived-Growth-Factor (PDGF) is known to have a chemotactic effect on PDL cells, and to promote collagen synthesis (Matsuda et al., 1992). When combined with IGF-1 or dexamethasone synergistic effects on PDL cell mitogenesis were observed (Matsuda et al., 1992 ; Rutherford et al., 1992a, 1994). Recently it was suggested that the chemotactic and mitogenic response of PDL fibroblasts to PDGF was suppressed by P. gingivalis by down-regulation of PDGF cell surface receptors (Matsuda et al., 1996).
Transforming-Growth-Factor-beta (TGF-β1) strongly enhances the production of extracellular matrix in many cell types and also in PDL cells (Matsuda et al., 1992). It does not promote PDL cell migration and has modest effects on PDL cell mitogenesis (Matsuda et al., 1992 ; Oates et al., 1993 ; Dennison et al., 1994). TGF-inhibits epithelial cell proliferation (Lynch et al., 1989a ; Tucker et al., 1984).
Basic Fibroblast-Growth-Factor (bFGF) has been shown to have chemotactic and mitogenic effects on PDL cells (Terranova et al., 1989).
Insulin-Like-Growth-Factors-I and II (IGF-I and II) : IGF-I has been shown to be chemotactic for PDL derived cells and displayed strong mitogenic and protein synthesis enhancing effects, as well (Matsuda et al., 1992). IGF-I receptors were demonstrated on the surface of PDL cells (Blom et al., 1992). PDGF stimulation promoted IGF-I receptor expression in fibroblasts, thereby allowing greater effects of exogenous IGF-I (Rubini et al., 1994).
Cementum-Derived Growth Factor (CGF) appears to be only found in cementum (Narayanan and Yonemura, 1993). It is mitogenc for PDL cells and may be involved in the differentiation of cementoblasts (Yonemura et al ., 1992, 1993).
Bone Morphogenetic Proteins (BMPs) : BMP-7 (Osteogenic Protein-1 = OP-1) was not mitogenic for PDL cells (Rutherford et al., 1994), however changed their phenotype by stimulating their alkaline phosphatase activity in a dose and time dependent manner. BMP-2 and -12 did not show a mitogenic effect on PDL cells (Nguyen et al., 1995). Future studies exploring the effects of other BMPs on PDL cells would be of great interest.
An abundance of growth factors are stored in bone matrix and are thought to be responsible for the coupling of bone formation and resorption (Baylink et al., 1990, 1993), possibly by delayed paracrine mechanisms. During bone repair there is a temporal expression of multiple growth factor genes and their gene products (Bolander, 1992 ; Andrew et al., 1993a, 1993b ; Joyce et al., 1990). It is reasonable to believe that several interactions between various GF exist during bone formation and repair. Even though the role of polypeptide growth factors on bone formation has been studied extensively, there is insufficient information specifically on alveolar bone.
Platelet-Derived-Growth-Factor (PDGF) stimulates mitogenesis and chemotaxis as well as protein synthesis in bone organ culture (Graves et al., 1989 ; Hock and Canalis, 1994 ; Hughes et al., 1992). PDGF is released by human osteoblast like cells (Zhang et al., 1991).
Transforming-Growth-Factor-beta (TGF-β) : The effects of TGF-β1 seem to be dependent on the kind of bone cell, dose and local environment (Centrella et al., 1987 ; Sporn and Roberts, 1992). TGF-β is known to stimulate extracellular matrix synthesis (Bonewald and Mundy, 1990 ; Pfeilschifter et al., 1990) and to inhibit matrix degradation (Edwards et al., 1987 ; Overall et al., 1991). TGF-β1 genes are expressed during normal fracture healing (Andrew et al., 1993b).
Fibroblast-Growth-Factor (FGF) stimulates bone cell proliferation (Canalis et al., 1988) and angiogenesis (Folkman and Klagsbrun, 1987).
Insulin-Like-Growth-Factors-I and II (IGF-I and II) are found in large amounts in bone. IGF-II is the most abundant growth factor in bone matrix (Mohan and Baylink, 1991a and b). IGF-I is produced by osteoblasts and promotes bone formation by stimulating differentiation, proliferation and collagen biosynthesis (Canalis, 1980 ; Canalis et al., 1988 ; Mohan et al., 1986). It is thought to regulate bone formation in an autocrine manner (McCarthy et al., 1989). IGF-II may be not as potent as IGF-I in enhancing bone formation (McCarthy et al., 1989). Synergistic effects of IGF-I and other growth factors (bFGF, PDGF, TGF-β) have been shown on osteoblast mitogenesis (Giannobile et al., 1997 ; Piche and Graves, 1989). There is evidence that IGF-I or IGF-II combined with other GFs may augment the osseous wound healing process (Pfeilschifter et al., 1990 ; Giannobile et al., 1997).
Bone Morphogenetic Proteins (BMPs) are structurally related members of the TGF-β superfamily (fig. 3) and are in contrast to other growth factors able to « induce » de novo bone formation (fig. 4) (Urist, 1965 ; Urist and Strates, 1971 ; Sampath and Reddi, 1981, 1983 ; Wozney et al., 1988 ; Luyten et al., 1989 ; Celeste et al., 1990 ; Sampath et al., 1990, 1992 ; Özkaynak et al., 1990, 1992). BMPs are abundant in bone and are produced by several cell types including osteoblasts (Urist et al., 1977). The main effects of BMPs are to commit undifferentiated pluripotential cells to differentiate into cartilage and bone-forming cells (Wozney, 1992 ; Reddi and Cunningham, 1993 ; Asahina et al., 1993 ; Knutsen et al., 1993). BMPs were shown to regulate growth factor gene expression in rat calvaria cells (Yeh et al., 1998). They may act synergistically with IGF-1 to stimulate osteoblastic cell differentiation and proliferation (Yeh et al., 1997). BMPs play a role in dentinogenesis and can promote reparative dentin formation in vivo (Jepsen et al., 1997).
In a pilot study in 3 beagle dogs Lynch et al. (1989b) evaluated the combination of PDGF/IGF-I in the naturally occuring periodontitis model. Test theeth, that had received 1 µg of the growth factor combination in 75 µl methylcellulose gel carrier, showed new bone formation and cement deposition as early as 2 weeks after flap surgery, whereas control teeth treated with the carrier alone exhibited a long junctional epithelium.
In a larger study in 13 beagle dogs with naturally periodontitis Lynch et al. (1991a) administered 3 µg recombinant human PDGF-B in combination with 3 µg recombinant human IGF-1 in 80 µl methylcellulose gel. After 2 or 5 weeks a significant 5-to 10-fold gain in bone and cementum was observed as compared to control (carrier alone). The half-life of the growth factors at the site of application was determined as well and was 3 h for IGF-1 and 4.2 h for PDGF-B. Less than 4 % of radiolabeled growth factors could be found after 96 hours. The authors speculated that a single application of exogenous growth factors had stimulated a cascade of wound healing events that would have continued even in their absence.
The same PDGF/IGF-1 combination was evaluated by Rutherford et al. (1992b) in a different animal model. In 3 monkeys with ligature-induced Porphyromonas gingivalis infected periodontitis approximately 50 % of new attachment had formed after 4 weeks of healing.
In a subsequent study utilizing the non human primate model Rutherford et al. (1993) evaluated a combination of rhPDGF-BB/dexamethasone in a bovine collagen carrier matrix. In comparison to control defects treated with carrier alone they found after 4 weeks 5 times more cementum and attachment formation and a 7 times higher bone regeneration.
Giannobile et al. (1994) compared the results of a single PDGF/IGF-I treatment of naturally occuring periodontitis in dogs with a similar treatment of experimental periodontitis in nonhuman primates. They found significant regeneration compared to control treatment for both animal models. Whereas dogs exhibited more bone regeneration, the nonhuman primates showed more new attachment.
In a subsequent study Giannobile et al. (1996) further investigated recombinant PDGF and IGF-I either in combination or each factor individually for periodontal regeneration in 10 cynomolgus monkeys. Their results could demonstrate that IGF-I alone did not significantly alter wound healing. However, PDGF-BB alone could significantly stimulate new attachment formation while the PDGF-BB/IGF-I combination resulted in significant increases in new attachment formation and osseous defect fill.
Wang et al. (1994) used autoradiographic methods to analyze the influence of PDGF on fibroblast proliferation in periodontal fenestration defects in 6 mongrel dogs.
Compared to control defects without PDGF (with or without ePTFE-membrane) PDGF treated defects (with or without ePTFE-membrane) showed a significantly enhanced fibroblast proliferation after one and after 7 days.
The combination of PDGF application with guided tissue regeneration (GTR) was further evaluated by Park et al. (1995) and Cho et al. (1995). They introduced the concept of « PDGF-modulated GTR-therapy (P-GTR) ». Recombinant human PDGF-BB solution was applied to root surfaces that had been demineralized by citric acid. According to the authors the advantage would be that PDGF bound to the root surface would be slowly released over a prolonged period of time. Subsequently the defect was covered by an ePTFE-membrane and a coronally positioned flap. This treatment modality was compared with GTR-therapy in 6 dogs with surgically created class II furcation defects. Special emphasis was placed on topical and systemic infection prophylaxis. Significantly more bone and new attachment was found with P-GTR after 8 and 11 weeks. When using PDGF-BB alone in a separate experiment, enhanced bone regeneration was found already after 5 weeks. However, this was often accompanied by ankylosis and therefore could not be recommended by the authors.
An FDA phase I/II clinical trial was performed in 2 centers to evaluate the safety and efficacy of the combination of rhPDGF-BB/rhIGF-I for periodontal regeneration (Howell et al., 1997). Thirty-eight patients with intrabony and furcation defects were treated in a split-mouth design. Two doses (50 µg/ml or 150 µg/ml of each factor) were evaluated and compared to open debridement or carrier alone. Re-entry procedures after 6 to 9 months revealed significant effects for the high dose treatment. In intrabony defects 2 mm of gain in vertical bone height and 42 % of osseous defect fill were observed. Furcation defects responded with 2.8 mm of horizontal defect fill. A wide panel of safety parameters was determined. No safety concerns were found. No patient developed antibodies to the recombinant proteins. The authors concluded that additional studies, possible using higher doses, are warranted to more fully characterize and optimize the effects of this growth factor combination on periodontal regeneration in humans.
Even though there now is convincing preclinical and clinical data available for PDGF/IGF-I growth factor therapy this approach is presently no longer pursued. The reasons are not known and may possibly be related to the process of marketing approval.
Only 2 studies have evaluated other growth factors for periodontal regeneration.
Selvig et al. (1994) applied a growth factor combination (IGF-II/bFGF/TGF-β1) in a collagen sponge in experimental fenestration defects in 4 beagle dogs. In contrast to the control treatment they found a delayed bone formation following growth factor therapy.
In a recent study, Wikesjö et al. (1998) evaluated a combination therapy of TGF-β1 and guided tissue regeneration in 5 beagle dogs. Recombinant TGF-β1 was delivered in a calcium carbonate/starch carrier protected by an ePTFE-membrane and a coronally postioned flap. Only limited bone and cementum formation was observed that did not differ from control treatment (carrier alone plus GTR). Based on these findings the authors concluded that this treatment would have limited clinical potential.
The implantation of demineralized freeze dried bone allograft (DFDBA) has a long tradition in periodontics. Since the early publications by Urist (1965, 1971) periodontists have tried to utilize the osteoinductive factors presumably present in the graft for the stimulation of periodontal bone regeneration. Indeed, BMP-2, -4 and -7 were found in commercially available bone preparations of different bone banks (Shigeyama et al., 1995). However, in contrast to fresh preparations the biological activity appeared to be reduced (Shigeyama et al., 1995) and the ostoinductive properties of different preparations showed a high variability (Schwartz et al ., 1996). Moreover, Becker et al. (1995) following their investigations on the osteoinductive properties of DFDBA questioned the rationale for commercially available demineralized bone in periodontics. They demanded the loading of a carrier matrix with recombinant BMPs of known quality and quantity.
Bowers et al. (1991), in the first and only clinical trial using BMP for periodontal regeneration in humans, evaluated the effect of osteogenin (BMP-3) extracted from human bone for the healing of intrabony periodontal defects. Osteogenin was delivered in a DFDBA carrier matrix. In 36 defects in 8 patients healing proceeded following removal of the crown in a submerged environment and in 50 defects in additional 6 patients in a transgingival fashion. Defects treated with either carrier matrix or with non osseous collagen served as controls. Block biopsies were obtained after 6 months and healing was histologically evaluated. Whereas in the submerged environment the combination of osteogenin/DFDBA was significantly superior to DFDBA, the observed differences did not reach statistical significance in the transgingival model, the clinically relevant situation. The least favourable results were obtained with the collagen matrix, with or without the osteogenin. No immunological reactions due to osteogenin were found.
Ripamonti et al. (1994) in a pilot study in 4 monkeys tested the effect of a BMP-extract (bovine bone extracts, containing mostly BMP-2 and BMP-3) in an insoluble collagenous bone matrix (ICBM) for healing of 8 surgically created deep mandibular class II furcation defects. Eight contralateral defects treated with the carrier material served as controls. After 2 months there was a significantly enhanced regeneration of cement, periodontal ligament and bone in BPM/ICBM treated furcations. Using partially purified bovine BPM incorporated in a fibrous collagen membrane, Kuboki et al. (1998) demonstrated periodontal regeneration in class II furcation defects in 3 monkeys after 12 weeks.
Sigurdsson et al. (1995) applied recombinant human BPM-2 in a carrier consisting of resorbable PLGA-micro-particles using the supraalveolar defect model in 6 beagle dogs. Reconstructive surgery included application of test substance on the test side and of the carrier on the control side. To facilitate protected healing crowns were cut and flaps were sutured above the teeth (submerged model). After 2 months a substantial regeneration of bone (and cementum) was observed in test defects, significantly superior to control treatment. The incidence of root resorption was less in test sites, the incidence of ankylosis was similar to control treatment. In a subsequent study by the same group substantial BMP-induced periodontal regeneration could also be observed in the transgingival model (Sigurdsson et al., 1996). Healing results were significantly influenced by the kind of carrier material (6 different carriers for BMP-2 evaluated) that was used.
Kinoshita et al. (1997) performed periodontal reconstructive surgery with BMP-2 in a gelatin and polylactic acid polyglycolide acid copolymer carrier in ligature induced circumferential periodontal defects in 6 beagle dogs. Histometric evaluation after 3 months demonstrated significantly more new bone and cementum formation with no signs of ankylosis as compared to carrier alone.
King et al. (1997) studied the effects of rhBMP-2 in a rat fenestration defect model. Following 10 days of healing significant bone formation and 100 % more cementum formation was noted as compared to controls. However, after 38 days complete healing was found on both sides, leading the authors to the conclusion that in this model BMP-2 would accelerate bone and cementum formation during early wound healing.
Ripamonti et al. (1996) evaluated the effects of rhBMP-7 (OP-1) on healing of class II mandibular furcation defects. A total of 6 defects in 3 animals received BMP-7 at a concentration of either 0.1 or 0.5 µg/mg collagen matrix carrier. No bone formation was observed, however substantial new cementum formation. The authors concluded that BMP-7 at the given concentrations stimulated the cementoblast phenotype.
Jepsen et al. (1998) demonstrated the possibility of substantial bone regeneration and new cementum formation in class II furcations of non-human primates by using higher concentrations of rhBMP-7 (2.5 µg/mg) (fig. 5, 5b and 6).
Giannobile et al. (1998) evaluated different concentrations of rhBMP-7/OP-1 in a dose study in 18 beagle dogs. At a dose of 7.5 µg/mg collagen carrier a significant stimulation for all wound healing parameters was found that was statistically different from either vehicle or surgery-alone sites. No significant increase in root ankylosis was found.
The formation of not only bone but also of a new attachment apparatus following administration of BMPs is difficult to explain. It can be speculated that following the initiation of the wound healing cascade by BMPs other cytokines and/or growth factors stimulate the differentiation of cells to other non-osseous periodontal phenotypes, since direct mitogenic effects of BMP on periodontal ligament cells appear unlikely (Rutherford et al., 1994). Future research, including BMP receptor studies in periodontal tissues, will hopefully help to better understand the molecular mechanisms of BMP modulated periodontal wound healing.
In summary, there is strong evidence from different pre-clinical models that rhBMP-2 and -7 can stimulate periodontal regeneration. Human clinical trials are in progress to determine the safety and efficacy of recombinant morphogenetic proteins for periodontal reconstruction with the first results being anticipated in the year 2000.
It has been shown that long-term success of any implant under function depends on the achievement of direct bony anchorage (Brånemark, 1983). Thus, the two most important objectives for the use of growth and differentiation factors in implant dentistry are to increase bone implant contact (BIC) and to achieve a faster osseous integration (Becker et al., 1992). Other aims may be to improve the quality of bone surrounding the implant and to augment deficient implant sites.
A mixture of platelet derived growth factor and insulin like growth factor-1 (PDGF/IGF-1) in a carboxymethyl-cellulose gel as a carrier was used in a canine study. The rough implant surfaces were coated with the gel. In edentulous recipient sites in the mandible a significant acceleration of osseous integration was demonstrated, however the study failed to demonstrate a significant difference in BIC after 21 days compared to the placebo group and the bare implant surface alone (Lynch et al., 1991b). The same group reported a canine study where implants were placed in fresh extraction sites with vestibular dehiscence defects using barrier membranes and growth factors. Again an acceleration of formation of BIC was observed for the PDGF/IGF-1 group, and additionally, an increase of BIC was observed (Becker et al., 1992). In an orthopedic setting a similar effect was observed for TGF-β (Lind et al., 1996).
A natural source of PDGF, platelet rich plasma (PRP), has been demonstrated to be useful to support the healing of autogenous bone grafts (Marx et al., 1998). However, at present no data on the use of PRP in conjunction with the insertion of dental implants are available.
Bone morphogenetic proteins (BMPs) have been reported to enhance osseous contact of dental implants. Some of these studies used a naturally-sourced bovine BMPs preparation in a canine mandibular site (Wang et al., 1993, 1994 ; Yan et al., 1994) but quantitative data were not reported. rhBMP-2 was used in an in vitro assay and stimulated osteoblastic cells on a titanium surface (Ong et al., 1997). In a canine study using rhOP-1 in combination with bone derived type I collagen in fresh extraction sites in the mandible with simultaneous insertion of dental implants, the osseous contact of HA-coated implants was measured to be 80 %, although the difference to the controls was not significant (Cook et al., 1995). However, in a recently published study using rhBMP-2 in collagen carrier for ridge augmentation of the canine mandible with implant insertion the contact rate in regenerated bone was only 29 % after 16 weeks (Sigurdsson et al., 1997). Hanisch et al. (1997a) could demonstrate the possibility of a re-osseintegration of dental implants in a non-human primate peri-implantitis model after the use of rhBMP-2 in a collagen sponge. After 4 months of healing a BIC of 29 % was found in the defect area. Within the area of new bone formation a BIC of 40 % was found in rhBMP-2 treated sites. In a sinus augmentation study in miniature pigs using rhOP-1 with a deproteinized bone xenograft as a carrier and simultaneous insertion of dental implants a significant difference in BIC between test and placebo group was found. BIC on the test side (rhOP-1 and deproteinized bone xenograft) was 80 % compared to 38 % in the control sites in the regenerated bone (placebo and deproteinized bone xenograft) (Terheyden et al., 1999) (fig. 7 and 8, 8b).
A large number of experimental studies have investigated bone to implant contact rates following surface modification without the use of BMPs. A 61 % BIC in mandibular bone is a representative value for a titanium implant (Arvidson et al., 1990). In canine maxillae BIC after 4 months ranged between 43 % for a machined and 65 % for a TiO2-blasted surface (Eriksson et al., 1994). In a miniature pig study different types of titanium implants were inserted in cancellous bone around the knee joint. Among the titanium surfaces after 6 weeks the highest percentage of BIC was observed in rough sandblasted titanium (58 %) significantly superior to plasma sprayed titanium (38 %) and electropolished titanium (25 %). An acid pretreatment (33 % vs. 58 %) and surface roughness (medium grit sandblasting 22 % vs. large grit 58 %) significantly increased BIC (Buser et al., 1991). Numerous attempts have been made to increase BIC by using hydroxyapatite (HA) coatings. Contact values of 71 % (HA-coating) compared to titanium plasma-sprayed implants (55 %) and machined surface (46 %) were demonstrated in the mandible of dogs after 3 months (Weinlaender et al., 1992). HA-coating enhanced bone contact in rabbit femora compared to plasma-sprayed implants (75.9 % vs 59 %) after 6 months (Gottlander et al., 1992). In a miniature pig study after 6 weeks HA-coated implants (BIC : 69 %) significantly surpassed different types of titanium surfaces (BIC : 21 % - 58 %) (Buser et al., 1991).
Although the studies are not easily compared due to different animal models, experimental periods and surface characteristics, none of the above studies using surface modifications achieved an osseous integration as high as the 80 % reported in two studies using rhOP-1 either in extraction sites (Cook et al., 1995) or in regenerated bone (Terheyden et al., 1999). Thus, it can be concluded that rhOP-1 enhanced osseous integration in preclinical studies.
BMPs have also been reported to accelerate bone formation around the implant. In the above cited canine study using a bovine BMP preparation extensive bone apposition onto the implant was observed after 4 weeks by light-and scanning electron microscopy (Wang et al., 1993 ; Yan et al., 1994 ; Wang et al., 1994). In a pilot study in monkeys using a purified bovine osteogenic protein on a collagen carrier, rapid bone apposition onto the implant was observed after 3 weeks (Rutherford et al., 1992c). In a minipig sinus augmentation study, deposition of calcified material occurred on the implant surface after 2-3 weeks on the rhOP-1 side and after 8-9 weeks in the controls as monitored by fluorochromic labelling (fig. 7) (Terheyden et al., 1999). In contrast, in a canine study comparing periimplant defects in the mandible treated with and without rhBMP-2 significant differences to the controls were noted by radiographic evaluation after 12 weeks, but not after 4 weeks (Cochran et al., 1997).
The use of growth factors suchs as PDGF or IGF has not been reported for sinus or ridge augmentation. However, 2 studies evaluated rhBMP-2 delivered in collagen sponges for maxillary sinus floor augmentation in goats (Kirker-Head et al., 1997) and in humans (Boyne et al., 1997), although implants were not inserted at the time of first surgery. In a primate study comparing rhOP-1/collagen carrier or anorganic bovine bone in the sinus no dental implants were installed (Margolin et al., 1998).
The simultaneous placement of dental implants together with the administration of osteoinductive proteins appears to be advantageous (Terheyden et al., 1999). An implant recipient site containing bone morphogenetic proteins may present an osteoinductive environment for osteoprogenitor cells. These cells interact with extra-cellular matrix and surfaces in their environment (Ripamonti and Reddi, 1994). Thus it may be speculated, although it has not been tested yet, that for implants placed secondarily (second stage surgery) to bone augmentation with BMP the BIC rates would not be enhanced. In fact, in a second stage implantation study using rhBMP-2 on a collagen carrier in sinus augmentation in primates the bone to implant contact was 41 % and similar in rhBMP-2 augmented, control sites and in native bone adjacent to the augmented sinus (Hanisch et al., 1997b).
No growth or differentiation factor has been reported useful for the improvement of the local bone quality for example in type IV bone, although hypothethically this seems to be an interesting field for future research.
Clinical studies are sparse and of preliminary character. Purified natural BMP was reported to support the healing of dental implants (Sailer and Kolb, 1994). In a preliminary multicenter study using rhBMP-2 for sinus augmentation inconsistent augmentative success was reported (Boyne et al., 1997). Only one out of three patients treated with a sinus augmentation with rhOP-1/type I collagen carrier showed enough bone formation for implant placement (Groeneveld, 1999).
The success of tissue regeneration by growth and differentiation factors depends on the large scale purification and production of these molecules, as well as on suitable delivery systems for these factors to their target cells.
Much research has been performed to find optimal carriers for BMP application. The development of suitable delivery systems presents an important step for clinical growth factor therapy. Although a carrier matrix is not a prerequisite for BMP induced bone formation (Wang et al., 1990) it presents multiple advantages (Reddi et al., 1997) by immobilizing the protein in the target area. The carrier matrix not only defines the shape of the resulting bone, but allows smaller amounts of BMP to be active by retaining it until induction has occurred. An ideal carrier should bind the active protein and protect it against unspecific proteolysis. It should be biocompatible, non-immunogenic and biodegradable and not interfere with the wound healing process (Kenley et al., 1993 ; Miyamoto et al., 1993). It should facilitate rapid vascular invasion (Hollinger and Chaudhari, 1992) to enable contact between progenitor cells and the rhBMP bound to the carrier. A bone collagen matrix is the natural carrier for BMP, however, when using organic xenogenic materials or bone allografts the risk of disease transmission can not be ruled out (Lindhe and Cortellini, 1997).
In this regard resorbable synthetic materials such as polymers or calcium phosphate ceramics might be advantageous. Such alternative synthetic delivery systems have been evaluated in various animal models (Yamazaki et al., 1988 ; Ripamonti et al., 1992a, 1992b ; Miyamoto et al., 1993 ; Sigurdsson et al., 1995 ; Kinoshita et al., 1997).
Sigurdsson et al. (1996) evaluated different candidate carriers for rhBMP-2 in a screening study in the supralveolar defect model in the beagle dog (among others : bovine deproteinised bone mineral, PLGA-microparticles, PLA-granules). They found distinct differences in the amount and quality of the induced bone and cementum dependent on the type of carrier that was utilized. None of the materials appeared to be ideal in all aspects.
Similar observations were made when different carriers for rhBMP-7/OP-1 were studied in the rat mandibular augmentation model. Statistically significant differences for the carriers (among others : bovine deproteinised bone mineral, β-tricalciumphosphate, pycogenic hydroxylapatite) were found with regard to bone density, height of augmentation and bone quality (Jepsen et al., 1997 ; Terheyden et al., 1997). Differences in the release kinetics of rhOP-1 from the different biomaterials could partly explain the observed differences (Jepsen et al., 1999).
These findings indicate that in the future different delivery systems could be used for different surgical indications. Whereas a soft material that quickly resorbs might be well suited for the fill of periodontal defects, a horizontal ridge or sinus augmentation might require a more rigid, slowly resobable material with higher mechanical stability.
In addition to the question of the ideal delivery system other problems remain to be solved : What is the biological and therapeutic significance of the existence of multiple forms of BMPs ? What is the optimal therapeutic dose ? Future research should investigate different doses as well as molecular combinations to develop an activity profile for the different members of the BMP-family. Finally and most important, to confirm the preclinical data in patients with periodontitis, human biopsies as well as the results from randomised controlled clinical studies are needed.
A shortcoming of current delivery methods of growth factors to periodontal wounds is the short half-life of factors at the target site. The use of DNA delivery systems could become an alternative technique for the application of proteins to the wound site. Thus, the goal of gene therapy would be an elevated and sustained growth factor supply (days instead of few hours) in the healing wound. The rationale for this approach is based on the observations that growth factors are expressed up to 14 days during tissue injury as demonstrated for PDGF, bFGF and BMP (Antoniades et al., 1991 ; Werner et al., 1992 ; Nakase et al., 1994).
A prerequisite is the successful transduction of appropriate target cells. The efficient delivery of genes into cells can either be done in vitro or in vivo. Ex vivo therapies require transgene expansion from a tissue specimen. In vivo gene therapy resulting in higher but transient gene expression has been performed using plasmid DNA to healing skin (Slama et al., 1995 ; Andree et al., 1994) and bone wounds (Fang et al., 1996). Recently methods of transducing wounds by the in vivo microseeding technique have been introduced (Winkler et al., 1995 ; Giannobile et al. 1998). In another approach it was recently reported that human gingival fibroblasts after transduction with a recombinant adenovirus containing the OP-1 gene produced active BMP-7 resulting in bone formation in vivo (Hill et al., 1999).
Much research remains to be done to optimize gene expresssion, maximize the number of transduced cells and to evaluate whether periodontal and periimplant wound healing can be enhanced by gene transfer.
A large number of studies, performed over the last ten years, has demonstrated the possibility of tissue regeneration by recombinant polypeptide growth factors and morphogenetic proteins. There is evidence for the promotion of periodontal wound healing and oral implant integration by recombinant PDGF alone or in combination with IGF-I as well as by morphogenetic proteins such BMP-2 and BMP-7 from multiple in vitro and preclinical trials. Provided human clinical trials, of which the first has been published and others are ongoing, will confirm these findings and growth factor therapies will have received approval by the health authorities, the therapeutic use of these potent biologics would certainly add to our regenerative clinical strategies. In addition, in the future the development of gene therapy may become a novel approach in growth factor therapy for tissue engineering in periodontics and oral implantology.
Demande de tirés à part :
Søren JEPSEN, DDS, MD, MS, PhD, Department of Periodontics, Kiel University, Arnold-Heller-Str. 16, 24105 Kiel - GERMANY.