Articles
Nicola U. ZITZMANN * Carlo P. MARINELLO **
*Clinic of Fixed and Removable
Prosthodontics and TMJ Disorders,
University of Basel,
Hebelstr. 3
CH-4056 Basel Switzerland
The application of a barrier membrane intended to prevent epithelium from growing into a periodontal bony defect is generally described as Guided Tissue Regeneration (GTR). When a barrier is utilized to regenerate an alveolar bony defect in an edentulous area, the term Guided Bone Regeneration (GBR) is used. For both indications, the oral epithelium with its high turnover rate of 6-12 days should be excluded from the defect. However, while the aim of GTR is to...
Several grafting materials are available for Guided Bone Regeneration (GBR). Nevertheless, autogenous bone is still the first choice. The aim of this overview is : 1) to list the main advantages and disadvantages of the most common products and to describe their origins and some specific properties ; 2) to explain the indication for GBR treatment in cases where there are bone defects around implants ; 3) to describe the treatment in terms of sinus elevation techniques. Clinical studies utilizing the xenogenic material Bio-Oss® in combination with the resorbable collagen membrane Bio-Gide® are summarized. The clinical relevance of this combination is discussed and recommendations for applications are given.
The application of a barrier membrane intended to prevent epithelium from growing into a periodontal bony defect is generally described as Guided Tissue Regeneration (GTR). When a barrier is utilized to regenerate an alveolar bony defect in an edentulous area, the term Guided Bone Regeneration (GBR) is used. For both indications, the oral epithelium with its high turnover rate of 6-12 days should be excluded from the defect. However, while the aim of GTR is to regenerate a new attachment apparatus and component tissues, the goal of GBR is new bone formation. The requirements for an ideal membrane material are, on the one hand, sufficient rigidity so as to cover the defect like a tent and not to collapse into the defect ; it has to be stable enough for a period of about 4 months until the bone matrix has matured. On the other hand, the density of the material must be such that it allows sufficient nutrition of the surrounding soft tissues so as not to provoke wound dehiscences. In addition, the material has to be biocompatible so as not to cause an inflammatory reaction even if it is exposed to the oral cavity. The use of non-resorbable polytetrafluorethylene membranes (Gore-Tex®, Implant Innovation, West Palm Beach, FL) has been extensively described in the literature. Dahlin et al. (1989) augmented defect sites in rabbit tibia with e-PTFE-membranes (Gore-Tex®) and found significantly increased bone formation compared to the control sites. Several experimental investigations and clinical studies corroborate these results (Becker et al., 1990b ; Dahlin et al., 1995 ; Becker et al., 1994b). Buser et al. (1993) augmented the edentulous alveolar ridge prior to implant placement with e-PTFE-membranes. He used supporting screws and/or autogenous bone grafts as space-maintaining devices. By reinforcing the membrane, it was possible to reduce the risk of it collapsing into the defect (Schenk et al., 1994 ; Jovanovic and Nevins, 1995). However, the stiffer the membrane the higher the risk of membrane exposure due to reduced nourishment of the overlying soft tissues. Several authors reported an inflammatory reaction of the surrounding soft tissues and a reduced amount of bone fill when an e-PTFE-membrane was exposed to the oral cavity (Becker et al., 1994b ; Jovanovic et al., 1992). Plaque accumulation combined with bacterial infiltration of the membrane material was observed (Simion et al., 1994a). An additional surgical intervention was even necessary, in some instances, for membrane removal. Fugazzotto (1998) found that the majority of non-successful ridge augmentations (10/11) were associated with membrane exposure. Alternatively, resorbable membranes can be employed without any extra surgical step even in the event of a wound dehiscence (Sevor et al., 1993 ; Pajarola et al., 1994 ; Zitzmann et al., 1997a). However, since these biodegradable membranes are not sufficiently rigid, the defects must be filled with a « space maker » material. The use of a grafting material has the advantage of supporting the membrane and reducing the volume of the blood coagulum. Schulte (1964) has shown that the serum excretion, which occurs during the first 90 minutes after formation of the blood-clot, can be reduced from the normal 50 % vol. to 1.5 % of the entire volume when macerate spongiosa or other spongy materials are applied. As a result, the shrinkage due to fibril aggregation is reduced and the space between the surrounding bone walls and the coagulum decreased, thereby ensuring undisturbed replacement of the coagulum by granulation tissue. It is assumed that the latter is generally replaced by connective tissue containing mesenchymal cells, which have the ability to form osteoblasts and osteoclasts. Amler et al. (1960, 1969) found that osteoid appeared at 7 days in human extraction sockets, and that at least two thirds of the extraction fundus was filled with bone trabeculae after almost 6 weeks.
At the outset of the era when modern implant procedures were first used, implant position was determined by purely surgical considerations concerning the amount of available bone in a given location. Sharp residual bony ridges were reduced to create a 5-6 mm wide plateau (Adell and Lekholm, 1985). More recently, the approach has been primarily dictated by the prosthetic reconstruction planned and to a lesser extent by the presenting anatomical situation. With the proposed superstructure in mind, an adequate number of implants in ideal positions should be selected by means of « backwards planning ». However, despite research, we still know little about ideal implant lengths, and still do not know how much bone to implant contact is sufficient and how many implants are required for particular superstructures.
Several factors may influence implant longevity. Firstly, factors to do with the given bone morphology with its corresponding bone quantity and quality. Secondly, factors concerning a patient's individual features, such as expected loads, parafunctional habits and his/her oral hygiene ability. As a general rule, the entire available bone is used and the longest possible implant with bicortical anchorage or bucco-lingual stabilization is placed. Care should, however, be taken to preserve the neighboring anatomical structures (nerves, sinus maxillaries, floor of the nose) and to maintain the prosthetically driven implant position and angulation. This means that the available bone in the area affected should be estimated clinically by palpation of the alveolar ridge and with the help of the radiologic diagnosis. The practitioner evaluates whether an implant of adequate length can be placed in the site so as to ensure that primary stability, which is one of the pre-conditions for successful implant osseointegration, can be achieved. Depending on the bone quality, implants should be stabilized in at least 4 mm of existing bone (Lazzara, 1989 ; Becker and Becker, 1990a). If it has to be assumed that primary stability of the implant cannot be attained, then it will be necessary to consider a vertical and/or horizontal augmentation of the alveolar ridge as a first stage procedure (table I). In the lateral maxilla, the main cause of a reduction in vertical bone height is often an increase in pneumatisation of the maxillary sinus. This is, presumably, due to the pressure in the sinus cavity being higher than the atmospheric pressure. Instead of augmenting the area of the alveolar ridge crest, a sinus floor elevation and augmentation in an apical direction may be necessary in these cases (see chapter « Augmentation procedure »).
Another indication for bone regeneration is given when a portion of the implant remains exposed during placement. The exposed implant surface appears as a dehiscence defect around the implant shoulder or as a fenestration along the implant screw (fig. 1). The latter defect morphology is found predominately in the maxilla and often presents with buccal concavities due to a post-extraction collapse of the buccal alveolar bone enhanced by pressure from lip and the cheek. Dehiscence defects may occur in immediate or short-term delayed implant placement due to the incongruity between the cross section of the former root and the implant diameter (see chapter « Clinical considerations for GBR treatment »). In long-term delayed implant placement, dehiscence defects are often created when placing implants within thin, sharp crests (Zitzmann et al., 1997a). It is a point of controversy whether small defects should be treated with bone-regenerating techniques, especially in the light of the fact that bone resorption can be expected up to the first thread of screw-typed implants (Brånemark-type) within the first year of loading (Adell et al., 1981). It might even be preferable to « anticipate » this remodeling process by placing the implant shoulder supracrestally instead of placing a shorter implant. However, this concept should only be considered in areas of limited bone height, where there is sufficient intermaxillary space and reduced aesthetic demands. Dietrich et al. (1993) reported on an evaluation of IMZ implants where a thin buccal bone plate (≤ 1 mm) resulted in a reduced success rate compared to a bone thickness of more than 1 mm (89.3 and 96.6 % after 5 years, respectively). The authors suggested that a missing buccal bone plate would compromise an implant's long-term prognosis.
The importance of complete implant integration in bone can be explained, in summary, as follows :
Firstly, from a functional point of view, the goal is to ensure that load distribution over the surrounding bone is along the entire implant surface. Secondly, from an aesthetic point of view, the soft tissues need to be supported by the underlying bone. And thirdly, patient discomfort may arises because the periosteum is in too intimate contact with the exposed implant surface and is thereby scratching during movement of the overlying soft tissues.
Autogenous bone is the filler material of first choice. When transplanted immediately, autogenous grafts maintain their osteogenic potency and promote osteoinduction through the remaining vital cells (Boyne, 1991 ; for review see Tolman, 1995). Hence, the preserved osteo-progenitor cells and bone morphogenic proteins are supposed to induce the mesenchymal cells of the recipient bed to form new bone (Reddi et al., 1977). As far as the revascularisation and resorption of autogenous grafts are concerned, one has to distinguish between intramembraneous and endochondral ossifying donor sites. The latter comprise bone harvested from the iliac crest or the rib and shows less rapid development of the vascular bed. This means that a 6 months healing period is required before implant placement. Bone harvested from the calvaria, the tuberosity, the chin or the retromolar region of the mandible is of intramembraneous origin. It produces less resorption and allows implant placement at 4 months (Williamson, 1996). Grafting material taken from the iliac crest is associated with donor site morbidity, general anesthesia, hospitalization, postoperative pain and temporary immobility. Using autogenous bone taken from intraoral sources may reduce these disadvantages. However, there are patients who refuse to have bone harvested from the chin, the ascending ramus or the tuberosity. Thus, the use of a bone substitute has to be considered in order not to violate an intact area as a donor site. An alternative to autogenous grafts are allografts, xenografts or synthetic materials (table II, fig. 2).
Allografts are taken from cadavers (i.e. Dembone®, Pacific Coast Tissue Bank, LA, CA). They are defatted in ethanol and ether, sterilized by ethylenoxid or gamma irradiation and then freeze-dried. Hence, the products can be stored and transported at room temperature. During the drying procedure, alterations in protein configuration occur, which are associated with a decrease in antigenicity of the allografts. Bone samples can also be demineralized (DFDB) by hydrochloric acid, divesting the bone of inorganic salts and permitting the organic matrix to express its biologic potency (Urist and Strates, 1971 ; Reddi et al., 1987). Thereby, bone morphogenic proteins are released which will probably be able to initiate osteoinduction. This means that bone tissue differentiation is initiated by interaction with mesenchymal cells (Urist and Strates, 1971 ; Small et al., 1993). Mellonig et al. (1981a, b) prepared bone allografts from pigs prior to implanting the material in calvaria defects of the same species. The authors found a significantly higher percentage of new bone when DFDBA was used compared to FDBA, autogenous osseous coagulum, or autogenous bone blend. They concluded that DFDBA is a material of high osteogenic potential. Using DFDBA to regenerate bone around implants or to fill the sinus cavity after sinus floor elevation did not, in studies with dogs, demonstrate favorable results (Becker et al., 1992 ; Becker et al., 1995b ; Wetzel et al., 1995). However, in these studies fundamental biological principles were not respected inasmuch as commercially available products prepared from human cadavers were used. Since Reddi et al. (1987) showed that the ability of demineralized bone matrix to induce bone formation is species-specific, grafting material that is derived from humans and applied in dogs obviously cannot become integrated into bone. The preserved bone inductive proteins (osteogenins) may cause foreign-body reactions. Using DFDBA derived from the same species, Becker et al. (1995a) were able to reveal direct bone apposition at the particle surface in the light microscope. Tissue banks commonly screen donors by serologic testing for HIV, Hepatitis B and C virus, HTLV-I transmitting T-cell leukemia/lymphoma and syphilis (Malinin, 1992). Antigen-DNA is detected using ELISA tests or radioimmunoassays to rule out bacterial or virus-infected bone samples. As soon as the immunoglobulin G appears, viral antigens are rarely detectable, if at all. Therefore, both tests for antigen and antibody are necessary and appear to be specific. The specific enzymes are able to detect the respective antigen even when mutations have occurred. However, clearly only those antigens which are already known can be detected.
Xenogenic materials must be completely deproteinized due to their foreign sources. They are then in the form of inorganic matrices with an osteoconductive potential. The bioinert scaffold is conducive to bone growth and allows bone apposition from existing bone. Bone of bovine origin such as Bio-Oss® (Geistlich Biomaterials, Wohlhusen, Switzerland) or OsteoGraph/N® (CeraMed, Ultimatics, Inc., Springdale, AR) has been intensively investigated because it has a porous architecture similar to human bone (Klinge et al., 1992 ; Wetzel et al., 1995 ; Wallace et al., 1996 ; Zitzmann et al., 1997a ; Berglundh and Lindhe, 1997 ; Hämmerle et al., 1998). Bio-Oss® is deproteinized during heat processing at 300 °C for more than 15 hours, so that all organic and possible antigenic material is eliminated, but the crystalline structure is left intact (fig. 3a and 3b). After alkaline treatment, the material (consisting of hydroxyapatite and carbonate) is sterilized at 160 °C. Bio-Oss® is available as a cancellous block, cancellous or cortical granules, or cancellous granules combined with collagen of porcine origin. OsteoGraph/N® consists of bovine bone which has been heated to approximately 1 100 °C. At this high temperature, HA-ceramic with a large crystal size develops, the surface of the particles melts and loses its porosity (fig. 4a). Wallace et al. (1996), who used OsteoGraph/N-300® and N-700® in combination with autogenous bone, assumed that the material was completely resorbed after a 20-month period. In this case-report, the grafting material was replaced by maturing bone and a fatty, relatively avascular bone marrow.
Xenogenic materials can also be derived from corals (Small et al., 1993 ; Smiler et al., 1992 ; Jensen et al., 1996 ; Moy et al., 1993 ; Wheeler et al., 1996) (Interpore®, Interpore International, Irvine, CA) or algae (Ackermann et al., 1994) (Algipore®, Friatec, Mannheim, Germany). Interpore® is a non-resorbable carbonate consisting of hydroxyapatite with large crystal sizes (fig. 4b). Algipore® is a carbonate of phycogenic origin which is dissolved when placed in bone defects (fig. 4c). Hydroxyapatite (HA) bone substitutes, such as OsteoGraph/LD/D® (CeraMed, Ultimatics, Inc., Springdale, AR) or OsteoGen® (Stryker Dental Implants, Kalamazoo, MI), can also be synthetically produced by sintering calciumphosphate at more than 1 000 °C (Smiler et al., 1992 ; Jensen et al., 1996). OsteoGen® and Osteograph/LD® are materials of low density that slowly dissolve in the surrounding fluid, while the material of higher density (OsteoGraph/D®) is non-resorbable (fig. 4d). These materials should provide a scaffolding for the deposition of new bone and are thus considered as osteoconductive.
Other synthetic materials are resorbable tri-calciumphosphates such as Cerasorb® (Curasan, Kleinostheim, Germany ; fig. 4e ) or BioBase® (Pace Medical, Medisave, Freiburg, Germany ; fig. 4f ). Buser et al. (1998) used the tri-calciumphosphate Ceros® (Mathys AG, Bettlach, Switzerland) in combination with e-PTFE-membranes in bone defects in pigs. The authors showed a significantly higher percentage of bone fill after 6 months of healing with this combination than with other filler materials, such as Interpore® or DFDBA.
Bio-Gran® (Implant Innovation, West Palm Beach) is also synthetically fabricated and consists of glass particles with almost no crystalline structure (fig. 4g). It is postulated that its ionic cohesion results in osteoinductive properties (Schepers and Ducheyne, 1997).
When deciding upon the extraction timepoint for a tooth, the clinician should also take into account the different options for implant treatment, if this is considered. An immediate implant (IIP) is placed simultaneously with the extraction procedure. It is frequently applied in the upper anterior for single tooth replacement to shorten treatment time (table III). A short-term delayed implant placement (STDIP) is performed 6-8 weeks after extraction so as to allow the soft tissue to heal and thereby ensure a tensionless adaptation of the wound margins after implant insertion. When implants are placed at least 6 months after extraction, a so-called long-term delayed implant placement (LTDIP) is performed providing the extraction socket has healed (Zitzmann and Schärer, 1997c).
If implant surfaces are exposed after placement, the defect morphology has to be assessed and should influence GBR procedures. According to Zitzmann et al. (1997a), the defect morphology can be described as a 1-wall defect when less than 33 % of surrounding bone surfaces are available. This type of morphology presents the most difficult situation to treat as the nutrition as well as the stabilization of the grafting material is provided by only one bony wall. Fixation of the membrane is recommended for this defect morphology. Two-wall defects with 33-67 % of surrounding bone surfaces are distinguished from infrabony or 3-wall defects with more than 67 % of surrounding bone surfaces. The latter include funnel-like defect morphologies, which presents the most favorable situation for bone regeneration (see chapter « Results and consequences for the clinical appliance of GBR).
In general, any bone defect or augmentation site should be free of connective tissue and can, especially in cases of very dense cortical bone quality, be perforated with a small round bur to allow bleeding from the marrow compartments. Patients' own bone morphogenic proteins may, therefore, be accumulated underneath the membrane. The grafting material should be moistened with sterile sodium solution and then densely packed into the defect. The addition of Tetracycline powder (30 mg/0.5 g Bio-Oss®) which causes a yellow staining of the particles clinically, produces a local antibacterial reaction. An enhancement of the bone regenerative capacity has been discussed in literature (Drury and Yukna, 1991). However, no clinical evidence for such an effect has been given so far. Therefore, it seems that the addition of Tetracycline is not a requirement, and in any case should only be used in low doses so as not to entail a significant pH
-shift. Depending on the expected loads and on the morphology of the defect or the area to be augmented, the selected membrane should be fixed by additional nails or tacks (e.g. Frios®, Friatec, Mannheim, Germany). The following treatment sequence is recommended :
1. The cortical bone plate surrounding the defect is perforated with a small bur to create several holes (fig. 5a).
2. The soft tissue flap is checked for sufficient mobilization, and a periosteal release is performed, if necessary.
3. The defect size is measured and the membrane is adapted to the required size, allowing an overlap of the surrounding bone margins of about 3 mm. Placing the membrane in contact with the adjacent teeth is feasible so long as resorbable barriers are used, but this needs to be avoided with non-resorbable membranes.
4. The membrane is placed on the buccal bone plate and optionally fixed with 2 nails (making sure not to damage adjacent roots). If titanium pins are used, preparation of a hole is not necessary. Care must be taken to hold the insertion instrument perpendicular to the bony plate (fig. 5b).
5. The membrane is reflected to the buccal side and the defect is filled with the selected grafting material (e.g. Bio-Oss®). This should first have been moistened with sterile saline solution. The granules can be densely packed and excess saline solution is absorbed with a sterile gaze (fig. 5c).
6. The membrane is deflected to the lingual side and placed underneath the mobilized lingual flap. Deep horizontal mattress sutures can be applied, especially in the upper jaw, to fix the membrane at the inner surface of the lingual keratinized mucosa (fig. 5d).
7. Wound closure is performed with horizontal mattress sutures and additional criss-crossed single sutures.
Sinus elevations can be performed either laterally or from a caudal direction (fig. 6). When utilizing the latter with a crestal approach from the alveolar ridge, hand instruments are used to prepare the implant site in the planned position. The osteotomes (Implant Innovation, West Palm Beach) are tapped in a cranial direction pushing ahead of them the compressed medullary bone and, at the same time, carefully releasing the Schneiderian membrane from the sinus-floor. This method was first described by Tatum (1986) and modified by Summers (1994), who added bone-grafting material to fill out the space under-neath the membrane. The existing bone is preserved as no drilling takes place, with the exception of an initial perforation of the cortical bone with a round bur. Care must be taken not to displace the instruments buccally when thick palatal cortical bone is present (fig. 7). In these situations, adequate drilling is recommended in combination with the osteotomes. When a small perforation of the Schneiderian membrane occurs, a shorter implant should be used. Its length should be the same as the initially measured available bone. Of course, no grafting material is added and antibiotic coverage (e.g. Amoxicillin 375 mg tid.) should be prescribed for 10 days.
Performing a sinus elevation through an opening in the lateral wall of the sinus (antrostomy) was first described by Boyne and James (1980). The lateral window in the bone is made over the uppermost point of the residual alveolar ridge with a diameter of approximately 10 mm. A CT-scan is highly recommended to pinpoint the exact location of the sinus floor in the transversal sections of the Denta-Scan (Philips, Eindhoven, Netherlands) (fig. 8a). The Schneiderian membrane is carefully released and the loosened bony plate is cranially tapped into the sinus like a trap-door (fig. 8b). The membrane should be set free along the entire floor up to the medial bony wall of the sinus cavity. Generally, when the membrane is released from the bone, it tends to constrict, thus creating a space, which is then filled with bone-grafting material. When a septum in the maxillary sinus hinders the inward movement of the bone plate, then its' complete removal is indicated. After this the sinus membrane is released, the sinus filled and the bony covering repositioned. Depending on the residual bone height and the given bone quality, implants are placed simultaneously or in a second-stage procedure. In general, a minimum of 4 mm is required for the one-step procedure to attain primary stability of the inserted implant. Abutment connection is then carried out 6-9 months later. The two-step procedure is indicated when less than 4 mm bone height is present or when bone quality does not allow primary stability of the implant. Hence, implant placement should be performed 8-9 months after sinus elevation.
The use of non-resorbable membranes for GBR is increasingly being replaced by biodegradable barriers in combination with grafting materials even though good results are obtained with the former when uneventful healing occurs. Utilizing resorbable membranes reduces the risk of infection despite exposure of the membrane in the oral cavity. Generally, wound dehiscence occurs more often in immediate or short-term delayed implant placements due to the soft tissue deficit ; however, these timepoints have several other advantages. In their investigation of the factors that might influence the bone fill after GBR treatment with Bio-Oss® and Bio-Gide®, Zitzmann et al. (1999) found that immediate and short-term delayed implant placements showed better results (92 % bone fill) than the long-term delayed placements (80 % bone fill). The more advantageous 2-and 3-wall defect morphologies occurred more frequently with the earlier placement timings.
A clinical study investigating the osteotome technique has shown that an increase of 3.5 mm (average) in bone height can be achieved (Zitzmann and Schärer, 1998a). The technique was recommended for alveolar ridges with at least 6 mm residual bone height so as to allow a 10 mm implant (Brånemark-type) to be placed. Implants of less than 10 mm length should only be inserted if it is possible to splint them with neighboring fixtures of longer length. The lateral antrostomy is indicated when less than 6 mm residual bone height is present in the affected site or an increase of more than 4 mm is the goal.
Two-stage augmentation procedures, such as ridge augmentation and 2-staged lateral antrostomy, allow biopsies to be taken during the reentry. In a 22-year old woman, single tooth replacement in area 12 was planned, but implant placement was not feasible during initial surgery because the cortical bone plate was too thin (fig. 5a). The histologies shown in figure 9a and 9b present an area of the regenerated bone 6 months after ridge augmentation with Bio-Oss® (BO) spongiosa granules (size 0.25-1 mm). The grafting material is embed-ded in regenerated woven bone (RB), and the light microscope showed intimate contact between them. A row of osteoblasts between marrow compartments (MC) and woven bone can be identified. Blood vessels (Ve) within the marrow reflect its vitality without any histologic signs of an inflammatory reaction.
It has been shown in various studies (Zitzmann et al., 1997b, 1998b, 1999) that short-term delayed implant placement (STDIP) 6 weeks after tooth extraction is the most preferable placement timing as it combines the advantages of early and delayed insertion (table III). The capacity for bone regeneration seems to be greater with the early placement timings since functional loading by natural teeth is not absent for longer than 6 months. Amler et al. (1960, 1969) reported that complete epithelial closure of an extraction socket takes about 4-5 weeks, depending on the socket diameter. Hence, wound closure is feasible when applying STDIP. Several authors (Tallgren, 1972 ; Atwood, 1979 ; Carlsson and Persson, 1967a) reported that most resorption of the alveolar bone occurred during the first year, mainly during the first 2 months, after extraction. Carlsson et al. (1967b) observed that the labial bone plate was resorbed and partly replaced 6 weeks after extraction. However, when comparing immediate and short-term delayed implant placements, it appeared that both presented with similar defect sizes (mean 22.8 and 20.6 mm2, respectively), indicating that a 6-week waiting period did not cause any extensive bone loss (Zitzmann et al., 1999).
The requirements for bone-grafting materials can be summarized as follows. The grafting material has :
- to be biocompatible and non-allergenic ;
- to cause no antibody reaction ;
- to be radiopaque ;
- to have high compressive strength ;
- to be osteoconductive to allow for bone apposition in intimate contact.
Osteoinductive properties, such as those described for allogenic bone grafts, are critical because, as soon as protein structures are preserved, the material could possibly become antigenic, so that it is not possible to rule out any transfer of infectious diseases. The question as to whether the material should be resorbable and biodegrade during a specific time frame cannot be answered generally. It is known from autogenous bone grafts that their tendency to resorb can be reduced by functional loading when implants are placed in the grafted sites (Buser et al., 1995). With sinus grafts in sheep, Haas et al. (1998) observed an extensive resorption of the autogenous graft close to the elevated Schneiderian membrane. Schenk et al. (1994) have shown that regeneration underneath the e-PTFE-membrane starts with the deposition of woven bone, which is gradually replaced by lamellar bone. As known from fracture healing, the apposition rate of woven bone in the adult human is about 100 μm per day. Hence, regeneration of a defect 10 mm in diameter needs approximately 7 weeks, provided that the defect is surrounded by bone walls and regeneration starts from opposing walls. In the light of this, it is apparent that vertical augmentation with an absence of surrounding bone walls is very limited. Regeneration can only proceed from the bone base and there is no protection or stabilization of the graft. Even though Simion et al. (1994b) used a titanium reinforced e-PTFE-membrane and stabilized it with screws, they found a maximal vertical bone gain of 4 mm on average.
The question arises as to whether funnel-like defects around implants are in need of GBR treatment at all. Using e-PTFE-membranes to cover funnel-like defects around immediate implants in dogs, Becker et al. (1991) reported a significantly increased defect reduction compared to the control. On the other hand, it has been shown that such defect morphologies around immediate implants can be regenerated with autogenous bone only (Becker et al., 1994a). With a titanium implant surface on one side, ideally surrounded by intact bone walls, the morphology means that a 3 mm gap can possibly be bridged after 1 month, provided that no epithelial down-growth had occurred. However, the replacement of woven bone by lamellar bone takes several months due to the apposition rate of 1 μm per day (Frost, 1966). Evian et al. (1982) found alveolar trabecular bone being established with signs of remodeling in the healing socket 16 weeks after extraction in humans. Forming a Haversian system takes about 3 months (Frost, 1963). During this remodeling process, the primary scaffold is replaced by thicker trabeculae (Schenk et al., 1994) which reinforces and stabilizes the bone matrix. So, it is questionable whether an osteoconductive bone graft when included into the regenerated bone matrix should be resorbed at all, or whether resorption might cause a weakening of the regenerated bone substance. Once embedded into the mature bone matrix, the bone graft will be included and participate in the physiological remodeling process, which proceeds very slowly, especially when no additional special stimuli like implant placement or abutment connection have been introduced (Frost 1963 ; Berglundh and Lindhe, 1997) (fig. 9a and 9b). It can be concluded that the grafting material should have the potential for replacement by host bone ; however, with increasing defect size, a larger quantity of autogenous bone may be required.
Acknowledgement - The authors would like to thank Dr. Peter Schüpbach, PD (Department of Oral Microbiology and Immunology, University of Zurich) for performing the histologic sections and Renate Löffler (Department of Fixed and Removable Prosthodontics, University of Zurich) for her help in preparing the REM pictures. We would also like to acknowledge the assistance of several companies who provided samples of their products for the REM investigation.
Demande de tirés à part
: Nicola Ursula ZITZMANN, Assistant Professor Clinic of fixed and removable prosthodontics and TMJ disorders, University of Basel, Hebelstr.3, CH-4056 BASEL, Switzerland.