| Sign In to gain access to subscriptions and/or personal tools. |
The Influence of Barrier Membranes on Autologous Bone Grafts
1 Department of Oral and Maxillofacial Surgery, Correspondence: * corresponding author, p.f.m.gielkens{at}kchir.umcg.nl
In implant dentistry, there is continuing debate regarding whether a barrier membrane should be applied to cover autologous bone grafts in jaw augmentation. A membrane would prevent graft remodeling with resorption and enhance graft incorporation. We hypothesized that membrane coverage does not effect resorption and incorporation of autologous onlay bone grafts. We treated 192 male Sprague-Dawley rats. A 4.0-mm-diameter bone graft was harvested from the right mandibular angle and transplanted to the left. Poly(DL-lactide-  -caprolactone), collagen, and expanded polytetrafluoroethylene membranes were used to cover the grafts. The controls were left uncovered. Graft resorption at 2, 4, and 12 weeks was evaluated by post mortem microradiography and microCT. Analysis of the data showed no significant differences among the 4 groups. This demonstrates that the indication of barrier membrane use, to prevent bone remodeling with resorption and to enhance incorporation of autologous onlay bone grafts, is at least disputable.
Key Words: Bone resorption • bone transplantation • membrane, artificial • microradiography • micro-CT
Sufficient bone is necessary for the predictable osseointegration of dental implants and satisfactory aesthetics. The use of Guided Bone Regeneration Membranes has proven to promote bone regeneration in bony defects (McAllister and Haghighat, 2007). However, when a bone graft is applied to augment the jaw, there is continuing debate regarding whether a barrier membrane should be used to cover the augmented site (Chiapasco et al., 1999; Donos et al., 2002a). The bone graft serves as a scaffold and carrier for living cells. The barrier membrane is expected to prevent bone remodeling with resorption by shielding the graft from inhibiting factors and fibroblasts (Gordh et al., 1998), and by keeping the osteoinductive substances in situ (Linde et al., 1993; Zellin and Linde, 1997). This would enhance incorporation of the bone graft (Alberius et al., 1992) and improve the predictability of the augmentation (Donos et al., 2002b). Furthermore, a barrier membrane serves as a space-maintainer, allowing for bone regeneration in any remaining space (Antoun et al., 2001). Good clinical results with barrier membranes have been reported, and many clinicians cover bone grafts with a barrier membrane (Buser et al., 1996). However, membrane application increases costs (Chiapasco et al., 1999) and has a negative effect on Guided Bone Regeneration around dental implants in case of membrane exposure (Machtei, 2001). Moreover, the present best available evidence does not answer the question as to whether barrier membranes do prevent bone resorption in autologous onlay bone grafts (Gielkens et al., 2007). The objective of this study was to examine the preventive effects of 3 barrier membranes with regard to remodeling with resorption of autologous onlay bone grafts in the rat mandible. In addition, the effects of the membranes on graft incorporation were measured.
In the right mandibular angles of male Sprague-Dawley rats (Harlan Netherlands B.V., Horst, The Netherlands), a standardized 5.0-mm circular defect was drilled with a trephine (Hager & Meisinger GmbH, Neuss, Germany) (Kaban and Glowacki, 1981; Schortinghuis et al., 2003), and the obtained bone graft (4.0 mm diameter) was transplanted to the buccal side of the contralateral mandibular angle and fixed with a slowly degradable suture (Monocryl®, Ethicon, Johnson & Johnson, Amersfoort, The Netherlands) through a central 1-mm hole in the graft.
The rats were assigned to one of 4 groups: 3 membrane groups and 1 control group, in which no membrane was used. The membranes used were: (1) a copolymer sheet composed of 67–69% DL (15–85)-lactide and 31–33% One side of the PDLLCL membranes was rough. These membranes were applied with their rough side facing the bone to optimize integration and positioning. The wound was closed in layers with resorbable sutures (Vicryl Rapide 4-0, Ethicon, Johnson & Johnson, Amersfoort, The Netherlands). Postoperative pain relief was administered, and the diet was composed of standard laboratory food. After 2, 4, and 12 wks, rats were anesthetized by inhalation anesthesia and killed by an intracardial injection of pentobarbital, after which the mandibles were explanted and fixed in phosphate-buffered formaline solution. The study protocol was approved by the Animal Studies Review Committee, and was conducted in accordance with Institutional Guidelines (University Medical Center Groningen, The Netherlands).
Microradiography and Micro-computed Tomography Micro-computed tomography (microCT) images were obtained by means of a Siemens MicroCAT II pre-clinical cone-beam CT scanner (Siemens AG, Munich, Germany) (Gielkens et al., 2008). The CCD sensor measured 7 x 5 cm. The specimens were arranged in a three-dimensional array not exceeding field-of-view dimensions, to prevent truncation artefacts. The images obtained were in 3D, with an isotropic voxel size of 48 x 48 x 48 µm.
Measurement of Bone Modeling of the Graft and Graft Incorporation
The 3D microCT datasets were evaluated with image analysis software (AMIRA 4.1, Mercury Computer Systems, Chelmsford, MA, USA). A transverse plane rectangular to the mandible was chosen (Fig.  , B). Every third (corresponding to an interval of 144 µm) 2D microCT image between the proximal and distal limits of the bone graft was saved in a file. The outline of the graft was manually drawn, and the area was then calculated by image analysis software. For each sample, we calculated a mean volume of remaining graft by multiplying the sum of all marked areas by 144 µm (i.e., the thickness of 3 CT slides). In the same way, in every third microCT image, graft incorporation was measured, defined as a bony connection between the graft and the mandible. The percentage of incorporation was defined as the length of the incorporated part of the graft divided by the total length of the graft. Observations were scored on an ordinal scale, ranging from 1 to 4: When 0–25% of the graft was incorporated, a score of 1 was assigned, and scores of 2, 3, or 4 were assigned in cases of 26–50%, 51–75%, or 76–100% of incorporation, respectively.
Statistical analyses
The sample size that was determined based on the power analysis did not exceed the needed sample size of a defect experiment executed on the contralateral side in the same rat. Therefore, 48 rats per group were also used in the present study, yielding a total of 192 rats (mean weight 364 ± 17 g, range 320–407 g). Six rats died during surgery. In another 6 rats, the graft fractured during drilling, and in 3 samples the 1-mm drill hole, meant to be in the center, was located on the outline of the graft. These samples were excluded from statistical analysis, because the validity of volume calculations in these samples could not be guaranteed. In 2 specimens (PDLLCL and collagen at 12 wks), we could not identify the grafted bone with microCT. These 2 samples were also excluded. This resulted in a median group size of 15 (range, 13–16) for microradiography and a median group size of 11 (range, 7–11) for microCT. No wound infection or dehiscence occurred, and all other animals gained weight.
In all groups evaluated by microradiography, the mean ratio of graft and graft-surrounding area increased significantly from 2 to 12 wks (P < 0.05) (Table 1
Results from the present study indicate that the barrier membranes studied did not have a preventive effect on onlay bone graft resorption in the rat mandible, and that no positive effect of these membranes on graft incorporation is to be expected. Furthermore, there were no significant differences among the 3 membranes. The 2D perpendicular microradiographs indicate that the combined radiodensity of the mandibular bone and graft, in comparison with those of the graft-surrounding area, progressively decreased from 2 to 12 wks, suggesting nearly complete loss of volume of the graft bone at 12 wks in all groups. However, the decrease of radiodensity on the microradiographs can also be caused by a loss of volume of the mandibular bone under the graft, by a combination of both, or by a decrease of bone density. The 3D microCT measurements did not show a tendency of volume loss of graft bone, in any group. It must be noted that the 12-week group’s samples in particular showed an intact graft volume combined with a defect in the underlying mandibular host bone of approximately the size of the graft. The decrease in radiodensity as found by microradiography seems to be the result of volume loss of host bone and probably a decrease of the graft’s bone density. The defects in host bone might be caused by higher osteoclast activity in mandibular bone due to improved perfusion. Covered grafts, consisting of predominantly cortical bone, may be less susceptible to revascularization and might rely on previous host bone resorption both to become revascularized and to remodel (Salata et al., 2002). Some effect of the barrier membranes is suggested by an increase in bone volume within the first 4 wks as compared with that in the non-treated controls. However, the clinical relevance of this observation is unclear, since the differences in volumes are small, and the standard deviations are large. The smaller confidence intervals seen overall in the membrane-treated groups, with both microradiography and microCT, suggest a more predictable treatment outcome by membrane application. This is in agreement with previously reported results (Donos et al., 2002a,b). Conclusions from other studies were conflicting (Jensen et al., 1995; Chiapasco et al., 1999; Rasmusson et al., 1999; Antoun et al., 2001). A systematic review revealed that the best available evidence does not support membrane use (Gielkens et al., 2007). Furthermore, in a recently performed controlled trial with 31 individuals in each treatment arm, it was concluded that barrier membranes do not influence bone resorption (Meijndert et al., 2008). The results of microradiography are clinically more relevant, because the combined amount of bone in the graft area is represented. Another disadvantage of microCT was a lower-than-expected contrast. The outline of the graft was especially difficult to discriminate in the 12-week groups, where incorporation was almost complete. In this group, 2 cases were excluded for this reason, because the graft area could not be detected. As in other in vivo experiments (Chiapasco et al., 1999; Donos et al., 2002b), large standard deviations were seen. These may be partly related to this low contrast. A more plausible cause is the inter-individual variation. Graft diameter was similar in each specimen (4.0 mm diameter), but the height varied due to irregular anatomy and other inter-individual differences. A longitudinal study design with repetitive measurements of each sample in sedated animals could reduce standard deviations. Furthermore, this design would reduce the total number of animals or participants treated. Such a design was not yet available when the present study was approved, however. Another cause of the large standard deviations and disappointing success of the membrane treatment could be the method of fixing the grafts in the present study. It has been reported that rigid fixation of the graft is necessary for graft incorporation and the maintenance of graft volume (Raghoebar et al., 2006). For this reason, fixation with a micro-screw would have been preferred. However, titanium micro-screws would have interfered with the evaluation by microradiography and microCT. Also, a non-degradable material combined with a degradable membrane is not rational. Degradable micro-screws were considered to be too large to use in this study. Furthermore, favorable results for membrane treatment had been demonstrated previously when the graft was not fixed (Alberius et al., 1992; Gordh et al., 1998). An ideal barrier membrane is not yet available (Hardwick et al., 1994). Therefore, a new degradable barrier membrane (PDLLCL) (Meek et al., 2004) was compared with the standard non-synthetic degradable (collagen) and the standard synthetic non-degradable (ePTFE) reference materials. All membranes tested as equal compared with each other and to the control. PDLLCL has been shown to be biocompatible and non-cytotoxic (Meek et al., 2004), and the polymer is already applied in a commercially available nerve guide (Neurolac®, Polyganics, Groningen, The Netherlands) (Bertleff et al., 2005). PDLLCL has advantages when compared with the reference materials, because it is biodegradable and synthetic (Von Arx et al., 2002; Stavropoulos et al., 2004). Extrapolating results of animal research to the clinical (human) situation is always difficult. Loading of the onlay bone graft and membrane in the present study differs from the loading circumstances in transplanting to the alveolar process in human. However, the graft and membrane in the rat were subjected to loading to some extent by the overlying soft tissue. Moreover, this model has been recommended for use for graft coverage by barrier membranes (Donos et al., 2002a,b). In conclusion, this study shows that the indication of barrier-membrane use to prevent bone remodeling with resorption and enhance incorporation of autologous onlay bone grafts is debatable. No differences among the membranes were observed. It seems that barrier membranes are unnecessary in bone-grafting procedures. However, in this study, only onlay block grafts were used. When particulate bone is applied, a situation that is frequently seen in clinical practice (McAllister and Haghighat, 2007), the barrier membrane is necessary to secure these granules, but not to prevent bone resorption.
Mr. H. Bartels, Ms. Y. Heddema, and Mr. J de Vries are acknowledged for their assistance. Gratitude is expressed to Polyganics, Geistlich, and W.L.Gore & Associates for the provision of each of the membranes. This study is funded by the University Medical Center Groningen, The Netherlands. Received for publication June 25, 2007. Revision received July 9, 2008. Accepted for publication July 29, 2008.
Journal of Dental Research, Vol. 87, No. 11,
1048-1052 (2008)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
= 0.05.