Periodontal repair in dogs: gingival tissue occlusion, a critical requirement for GTR? Ulf M. E. Wikesjo ¨ 1 , Won Hee Lim 1 , Robert C. Thomson 2 and W. Ross Hardwick 2 1 Laboratory for Applied Periodontal and Craniofacial Regeneration, Department of Periodontology, Temple University School of Dentistry, Philadelphia, PA, USA; 2 Research and Development, Medical Products Division, W.L. Gore & Associates, Inc., Flagstaff, AZ, USA Wikesjo¨ UME, Lim WH, Thomson RC, Hardwick WR: Periodontal repair in dogs: gingival tissue occlusion, a critical requirement for GTR? J Clin Periodontol 2003; 30: 655–664. r Blackwell Munksgaard, 2003. Abstract Background: Design criteria for guided tissue regeneration (GTR) devices include biocompatibility, cell occlusion, space maintenance, tissue integration, and ease of use. Previous studies have established the importance of wound stabilization and space provision during the early healing sequel for successful GTR outcomes as well as evaluated biocompatibility, tissue integration, and clinical manageability of various biomaterials. The importance of cell or tissue occlusion has yet to be established. The objective of this study was to evaluate the role of tissue occlusion as a critical determinant for GTR outcomes. Methods: Routine, critical size, 5–6 mm, supra-alveolar, periodontal defects were created around the mandibular premolar teeth in six young adult Beagle dogs. Space- providing expanded polytetrafluoroethylene (ePTFE) membranes, with (macroporous) or without (occlusive) 300-mm laser-drilled pores, 0.8 mm apart, were implanted to provide for GTR. Treatments were randomly assigned to left and right jaw quadrants in subsequent animals. The gingival flaps were advanced to cover the membranes and sutured. The animals were euthanized at 8 weeks postsurgery for histologic and histometric analysis. Results: Three animals experienced wound failure within 2–3 weeks postsurgery resulting in exposure and removal of the occlusive ePTFE membranes. All defect sites, irrespective of membrane configuration or history of membrane exposure and removal, exhibited substantial evidence of periodontal regeneration including a functionally oriented periodontal ligament. To evaluate the biologic potential of GTR devices, only animals without wound failure and membrane removal were included. Alveolar bone regeneration for animals receiving occlusive and macroporous ePTFE membranes averaged (7SD) 3.271.1 versus 2.070.4 mm (p 5 0.3113). Cementum regeneration was enhanced in defect sites receiving the occlusive ePTFE membrane compared to the macroporous membrane (4.770.4 versus 2.370.2 mm; p 5 0.0167). Ankylosis was observed in one animal. Limited root resorption was observed in a second animal. Conclusion: Tissue occlusion does not appear to be a critical determinant for GTR. However, tissue occlusion may be a requirement for optimal GTR. Moreover, macroporous space-providing devices may increase the predictability of clinical GTR therapy. Key words: periodontal regeneration; tissue engineering; wound dehiscance; wound healing Accepted for publication 17 October 2002 Design criteria for guided tissue regene- ration (GTR) devices include biocompa- tibility, cell occlusion, space maintenance, tissue integration, and ease of use (Scan- tlebury 1993, Hardwick et al. 1995). Since the formulation of the principles of GTR (for a review, see Karring et al. 1993), studies have shown that the formation of a connective tissue attach- ment rather than a junctional epithelium onto the root surface following period- ontal reconstructive surgery is critically dependent on the stability of the wound (Wikesjo ¨ & Nilve ´us 1990, Wikesjo ¨ et al. 1991, Haney et al. 1993). If the gingival flap remains stabilized during the early healing period, the epithelium will not migrate onto the root surface. The use of a purpose-designed expanded polytetra- fluoroethylene (ePTFE) device has been demonstrated to contribute significantly to wound stability (Haney et al. 1993). Thus, it has been suggested that a junctional epithelium commonly observed following J Clin Periodontol 2003; 30: 655–664 Copyright r Blackwell Munksgaard 2003 Printed in Denmark. All rights reserved
Transcript
Periodontal repair in dogs:gingival tissue occlusion,a critical requirement for GTR?
Ulf M. E. Wikesjo1, Won Hee Lim1,Robert C. Thomson2 andW. Ross Hardwick21Laboratory for Applied Periodontal and
Craniofacial Regeneration, Department of
Periodontology, Temple University School of
Dentistry, Philadelphia, PA, USA; 2Research
and Development, Medical Products
Division, W.L. Gore & Associates, Inc.,
Flagstaff, AZ, USA
Wikesjo UME, Lim WH, Thomson RC, Hardwick WR: Periodontal repair in dogs:gingival tissue occlusion, a critical requirement for GTR? J Clin Periodontol 2003;30: 655–664. r Blackwell Munksgaard, 2003.
AbstractBackground: Design criteria for guided tissue regeneration (GTR) devices includebiocompatibility, cell occlusion, space maintenance, tissue integration, and ease ofuse. Previous studies have established the importance of wound stabilization and spaceprovision during the early healing sequel for successful GTR outcomes as well asevaluated biocompatibility, tissue integration, and clinical manageability of variousbiomaterials. The importance of cell or tissue occlusion has yet to be established. Theobjective of this study was to evaluate the role of tissue occlusion as a criticaldeterminant for GTR outcomes.
Methods: Routine, critical size, 5–6 mm, supra-alveolar, periodontal defects werecreated around the mandibular premolar teeth in six young adult Beagle dogs. Space-providing expanded polytetrafluoroethylene (ePTFE) membranes, with (macroporous)or without (occlusive) 300-mm laser-drilled pores, 0.8 mm apart, were implanted toprovide for GTR. Treatments were randomly assigned to left and right jaw quadrantsin subsequent animals. The gingival flaps were advanced to cover the membranes andsutured. The animals were euthanized at 8 weeks postsurgery for histologic andhistometric analysis.
Results: Three animals experienced wound failure within 2–3 weeks postsurgeryresulting in exposure and removal of the occlusive ePTFE membranes. All defectsites, irrespective of membrane configuration or history of membrane exposure andremoval, exhibited substantial evidence of periodontal regeneration including afunctionally oriented periodontal ligament. To evaluate the biologic potential of GTRdevices, only animals without wound failure and membrane removal were included.Alveolar bone regeneration for animals receiving occlusive and macroporous ePTFEmembranes averaged (7SD) 3.271.1 versus 2.070.4 mm (p5 0.3113). Cementumregeneration was enhanced in defect sites receiving the occlusive ePTFE membranecompared to the macroporous membrane (4.770.4 versus 2.370.2 mm; p5 0.0167).Ankylosis was observed in one animal. Limited root resorption was observed in asecond animal.
Conclusion: Tissue occlusion does not appear to be a critical determinant for GTR.However, tissue occlusion may be a requirement for optimal GTR. Moreover,macroporous space-providing devices may increase the predictability of clinical GTRtherapy.
Design criteria for guided tissue regene-ration (GTR) devices include biocompa-tibility, cell occlusion, space maintenance,tissue integration, and ease of use (Scan-tlebury 1993, Hardwick et al. 1995).Since the formulation of the principlesof GTR (for a review, see Karring et al.1993), studies have shown that the
formation of a connective tissue attach-ment rather than a junctional epitheliumonto the root surface following period-ontal reconstructive surgery is criticallydependent on the stability of the wound(Wikesjo & Nilveus 1990, Wikesjo et al.1991, Haney et al. 1993). If the gingivalflap remains stabilized during the early
healing period, the epithelium will notmigrate onto the root surface. The use ofa purpose-designed expanded polytetra-fluoroethylene (ePTFE) device has beendemonstrated to contribute significantly towound stability (Haney et al. 1993). Thus,it has been suggested that a junctionalepithelium commonly observed following
J Clin Periodontol 2003; 30: 655–664 Copyright r Blackwell Munksgaard 2003Printed in Denmark. All rights reserved
periodontal reconstructive surgery is asequel to wound failure (Wikesjo et al.1992, Wikesjo & Selvig 1999). Thegingival flap becomes detached from theroot surface during the early healingevents by mechanical means and/or dueto compromised adsorption/adhesion/ma-turation of the fibrin clot at the tooth–gingival flap interface, thereby permittingmigration of the epithelium into theinterface.
Studies have further shown thatalveolar bone and cementum regenera-tion is critically dependent on spaceprovision by the GTR device (Haney etal. 1993, Sigurdsson et al. 1994, 1995,Trombelli et al. 1999, Wikesjo et al.2002a). Limited or no regeneration ofalveolar bone and cementum has beenobserved in supra-alveolar periodontaldefects following gingival flap surgeryalone without space-providing GTRdevices, when the GTR device inadver-tently has collapsed or been compressedonto the root surface, and when thespace underneath the membrane hasbeen filled with a space-consumingslowly or nonresorbable biomaterial.Thus, also limited space provision doesnot appear conducive to regeneration ofalveolar bone and cementum. In thepresence of more extensive space provi-sion, cortical-appearing alveolar bonehas been observed filling the woundspace underneath an ePTFE devicewithin a 4-week observation interval(Haney et al. 1993). In the presence of‘‘abundant’’ space provision, the alveo-lar bone adopts a ‘‘physiologic’’ formalong the tooth surface (averaging 75%of 5-mm supra-alveolar periodontaldefects), leaving the remainder of thewound space underneath the membranefilled with dense fibrous connectivetissue (Sigurdsson et al. 1994, 1995).Cementum regeneration, as determinedby light microscopy, appears limitedwithin a 4–week healing intervalamounting to fractions of the defectheight, if at all appreciable (Haney et al.1993, Wikesjo et al. 2002a). Cementumregeneration following an 8-week heal-ing interval has been observed toaverage 40% of 5-mm supra-alveolarperiodontal defects in the presence ofwhat appears to be adequate spaceprovision (Sigurdsson et al. 1994,1995, Wikesjo et al. 2002b, c).
A recent study has evaluated theinfluence of ePTFE characteristics onwound healing, i.e., the influence ofePTFE porosity on osteogenesis (Zellin& Linde 1996). In a rat calvaria model,
ePTFE devices with a porosity of 20–25and 100mm increased the rate of osteo-genesis compared to less porous devices(o8mm). Increased permeability allow-ing transposition of extracellular matrixand cellular elements may have favoredosteogenesis. A concept of alveolar boneregeneration utilizing space provisionwithout connective tissue occlusionemerges from this observation. Thisconcept is also supported by observationsin a previous study evaluating the influ-ence of space provision without connec-tive tissue occlusion on osteogenesis(Karaki et al. 1984). Contralateral hor-izontal periodontal defects were surgicallycreated between the mandibular premolarteeth in dogs. A tissue-expanding goldmesh was applied on one side, while thecontralateral side served as surgical con-trol. Callus formation was enhanced indefects treated with the gold meshcompared to that in the surgical control.Evidently, osteogenesis in a periodontalenvironment may proceed in the presenceof space provision without strict occlusionof the gingival connective tissue.
The evidence presented has contrib-uted to a paradigm shift for GTR. First,wound stability is critical to maintainwound integrity for successful period-ontal regeneration (in part, woundstability may be provided by purpose-designed ePTFE devices). Second,space provision is critical for alveolarbone regeneration and for cementumregeneration. The importance of occlu-sion of cells from the gingival connec-tive tissue for alveolar bone andcementum regeneration has yet to beestablished. The objective of this pre-clinical study was to evaluate tissueocclusion as a critical design criterionfor GTR devices.
Material and Methods
Animals
Six male Beagle dogs (age 18–24months, approximate weight 15 kg) ex-hibiting intact mandibular premolardentition, without crowding or evidenceof periodontal disease, obtained from aUSDA approved dealer, were used.Animal selection and management,surgery protocol, and periodontal defectpreparation followed routines approvedby the Animal Care and Use Commit-tee, W.L. Gore & Associates, Inc.,Flagstaff, AZ, USA. The animals hadaccess to standard laboratory diet andwater until the beginning of the study.Oral prophylaxis was performed within
ePTFE, W.L.Gore & Associates Inc., Flagstaff, AZ,USA) were used (Fig. 1). The occlusivemembranes, custom-made and pre-shaped for the supra-alveolar, critical-sized, periodontal defect model (Wikes-jo et al. 1994), had a 15–25mm nominalpore size and were reinforced with alaminated polypropylene mesh. Thesemembrane characteristics have beenshown to support alveolar bone andcementum regeneration in the supra-alveolar periodontal defect model (Si-gurdsson et al. 1994). Membranes with-out provisions for tissue occlusionexhibited the same characteristics ex-cept for laser-etched 300-mm pores at0.8-mm (center to center) intervalsallowing for penetration of the gingivalconnective tissue. GORE-TEX ePTFEmaterials are considered to be inertbiocompatible materials as supportedby over 1 million clinical implants since1975.
Surgical protocol
Food was withheld the night beforesurgical procedures. The animals werepremedicated with atropine (0.02 mg/kgi.m.), buprenorphine (0.04 mg/kg i.m.),and flunixin meglumine (0.1 mg/kg i.v.).A prophylactic antibiotic (cefazolin;22 mg/kg i.v.) was administered. Gen-eral anesthesia was induced with diaze-pam (0.2 mg/kg i.v.) and ketamine(6 mg/kg i.v.). An endotracheal tubewas placed and the animals weremaintained on isoflurane gas (1–2%) in100% oxygen using positive pressureventilation. An i.v. line was placed andthe animals received a slow constantrate infusion of lactated Ringer’s solu-tion (10–20 ml/kg/h) to maintain hydra-tion while anesthetized. Routine dentalinfiltration anesthesia with epinephrinewas used at the surgical sites.
The maxillary first, second and thirdpremolar teeth were surgically ex-tracted, and the maxillary fourth pre-molars were reduced in height andexposed pulpal tissues sealed (Cavit
s
,ESPE, Seefeld/Oberbayern, Germany).This was done to alleviate the potentialof trauma from the maxillary teeth to
656 Wikesjo et al.
the experimental mandibular sites post-surgery.
Supra-alveolar, critical size period-ontal defects were created around thethird and fourth mandibular premolarteeth in the right and left jaw quadrantsin each animal (Fig. 1) (Wikesjo et al.1994). Briefly, buccal and lingual mu-coperiosteal flaps were reflected follow-ing buccal and lingual sulcular incisionsfrom the canine tooth to the secondmolar. The first and second premolarteeth, and the first molar were extracted.Alveolar bone was removed around thecircumference of the remaining premo-lar teeth using chisels and water-cooledrotating burs. The root surfaces wereinstrumented with curettes, chisels, andwater-cooled rotating diamonds to re-move the cementum. The crowns of theteeth were reduced to approximately2 mm coronal to the cemento-enameljunction (CEJ) and the cut surfacessmoothed. Exposed pulpal tissues weresealed (Cavit
s
, ESPE, Seefeld/Ober-bayern, Germany). The clinical defectheight, from the CEJ to the reducedalveolar crest, was set to 6 mm asmeasured with a periodontal probe.
Wound management
Occlusive and macroporous ePTFEdevices were alternated between the leftand right jaw quadrants in a split-mouthdesign in subsequent animals. To ensurean adequate blood clot underneath themembranes, 3 ml of autogenous venousblood was infused under the membrane.The blood was drawn using an i.v.catheter in aseptic manner and aspiratedblood was expelled underneath themembrane. The membranes were fixedto the reduced alveolar bone withmedical-grade stainless-steel tacks(FRIOS
s
Augmentation System, Fria-dent, Mannheim, Germany) designedfor these applications. A sham-surgerycontrol was not used since it has beenestablished that gingival flap surgeryalone will not provide conditions forperiodontal regeneration in the critical-sized supra-alveolar periodontal defectmodel (Sigurdsson et al. 1994, 1995).
Following membrane placement andclot positioning, the periostea werefenestrated at the base of the gingivalflaps to allow tension-free flap apposi-tion. The flaps were advanced, the flap
margins being adapted 3–4 mm coronalto the ePTFE membranes and sutured(GORE-TEXt Suture CV5, W.L. Gore& Associates Inc., Flagstaff, AZ, USA).Intrasurgery photographs were takenprior to and immediately after place-ment of the barrier membrane, andfollowing wound closure.
Postsurgery protocol
Animals were fed a canned soft dogfood diet the first 14 days postsurgery.Thereafter, the animals received stan-dard laboratory diet soaked in warmwater until thoroughly soft. The animalsreceived buprenorphine (0.04 mg/kgi.v., i.m., or s.q.) every 5 h for analgesiafor the first few days postsurgery. Abroad-spectrum antibiotic (enrofloxacin;2.5 mg/kg, i.m., b.i.d.) was used forinfection control for 14 days. Plaquecontrol was maintained by twice dailytopical application of chlorhexidine(chlorhexidine gluconate 20%, XttriumLaboratories, Inc., Chicago, IL, USA;40 ml of a 2% solution) until gingivalsuture removal and, thereafter, oncedaily (Monday through Friday) untilthe completion of study.
(a) (b)
(c) (d)
Fig. 1. Mandibular jaw quadrant (a) with surgically induced, critical-sized, discriminating, supra-alveolar periodontal defect (b). The defect isimplanted with a tissue-occlusive ePTFE membrane (c) or a macroporous ePTFE membrane (d) secured with stainless-steel tacks.Autologous blood is injected under the membrane to secure clot formation. The mucogingival flaps are then advanced to cover the teeth andmembrane devices and sutured. Healing interval: 8 weeks.
Periodontal repair in dogs 657
Radiographs were obtained immedi-ately postsurgery, and at 4 and 8 weeks.Gingival sutures were removed undersedation at approximately 10 days. Ob-servations of experimental sites withregard to gingival health, maintenanceof suture line closure, edema, andevidence of tissue necrosis or infectionwere made daily until suture removal, andat least twice weekly thereafter for theduration of the study. The animals wereanesthetized and euthanized at 8 weekspostsurgery. Following euthanasia, teethwith surrounding soft and hard tissuesand membranes were removed en bloc.
Histological processing and evaluation
The tissue blocks were fixed in 10%buffered formalin for 3–5 days, decalci-fied in 5% formic acid for 8–10 weeks,trimmed, dehydrated, and embedded inparaffin. Serial sections (7 mm) wereproduced in a buccal–lingual planethroughout the mesial–distal extensionof the teeth. Every 14th section wasstained with hematoxylin for observa-tions at 100 mm intervals.
The most central stained section ofeach root of the third and fourthpremolar tooth was identified by the sizeof the root canal. This section and theadjacent stained step-serial section oneither side were subjected to histometricanalysis. Thus, three subsequent step-serial sections, encompassing 0.2 mm ofthe mid-portion of the mesial and thedistal root for each premolar tooth, wereused for analysis. Analysis was per-formed using incandescent and polarizedlight microscopy (BX 60, OlympusAmerica, Inc., Melville, NY, USA), amicroscope digital camera system(DP10, Olympus America, Inc., Mel-ville, NY, USA), and a PC-based imageanalysis system (Image-Pro Plust, Med-ia Cybernetic, Silver Springs, MD, USA)by one experienced investigator maskedto the specific experimental conditions.The following measurements were re-corded for the buccal and the lingualtooth surfaces for each section:
� Defect height: distance between api-cal extension of the root planing andthe CEJ.
� Connective tissue repair: distancebetween apical extension of the rootplaning and the apical extension ofthe junctional epithelium.
� Cementum regeneration (height): dis-tance between apical extension of theroot planing and the coronal exten-sion of a continuous layer of newcementum or cementum-like depositon the planed root.
� Bone regeneration (height): distancebetween the apical extension of rootplaning and the coronal extension ofregenerated alveolar bone along theplaned root.
� Root resorption: combined linearheights of distinct resorption lacunaeon the planed root.
� Ankylosis: combined linear heights ofankylotic union between the regener-ated alveolar bone and the planed root.
Data analysis
Summary statistics (means7SD) basedon animal means for the experimentalconditions were calculated using theselected step-serial sections. Differences
(a) (b) (c) (e)
(d) (f)
Fig. 2. Photomicrographs of supra-alveolar periodontal defect implanted with the occlusive ePTFE membrane without wound failure,membrane exposure and removal. Overview (a & b; original magnification 2.5� , 4� , hematoxylin) and higher magnification (c, d; 8� ) ofthe buccal aspect of the defect show bone formation paralleled by regeneration of cementum and a functionally oriented periodontal ligament(e, f; 8� , polarized light). Animal #93.
658 Wikesjo et al.
between experimental conditions wereanalyzed using appropriate t-tests (N56).
ResultsClinical observations
Three jaw quadrants receiving theocclusive ePTFE membrane experi-enced membrane exposure within 1–2weeks postsurgery. These membraneswere removed within 2, 3, or 5 daysfollowing the observation of membraneexposure. Thus, the membranes wereremoved as early as 9, 11, and 21 dayspostsurgery. One macroporous ePTFEmembrane showed a small exposure at39 days postsurgery. This membraneexhibited a minimal local inflammatoryreaction and was maintained in situ forthe 8-week healing interval. Healingwas uneventful in the remaining ani-mals.
Radiographic observations
Radiographic evaluation at 8 weekspostsurgery showed significant newbone formation in jaw quadrants receiv-ing the occlusive ePTFE membranewithout exposure. Newly formed bone
filled the furcation area and approachedthe CEJ. Jaw quadrants with membraneexposure exhibited bone fill approxi-mating 30% of the defect height. Boneformation in jaw quadrants receivingthe macroporous ePTFE membraneapproximated 50–70% of the defectheight almost obturating the furcationarea in three animals. The newly formedbone did not appear ankylotic butcommonly exhibited a periodontal liga-ment space with lamina dura formation.
Histologic observations
Regeneration of periodontal structuresincluded newly formed bone approxi-mating 50–100% of the defect height insites receiving the occlusive ePTFEmembrane without wound failure(Fig. 2). The newly formed bone con-sisted of woven and lamellar bone withfibrovascular tissue in the marrowspaces. Cementum formation extendedabove the newly formed bone in allanimals approximating 86–100% of thedefect height. A functionally orientedperiodontal ligament with inserting fi-bers was observed at all sites. Ankylosis
was observed in one site. Healingcharacteristics at defect sites with ahistory of wound failure and earlymembrane removal were similar to thatobserved in defects without such heal-ing aberrations (Fig. 3).
Regeneration of alveolar bone andperiodontal attachment in defects re-ceiving the macroporous ePTFE mem-brane did not remarkably differ fromthat observed in defects receiving theocclusive ePTFE membrane (Figs. 4and 5). Newly formed bone consistingof woven and lamellar bone withfibrovascular marrow was observed.Cementum formation reached the levelof newly formed bone. Importantly, afunctionally oriented periodontal liga-ment with inserting fibers was observedat all sites. In sites exhibiting evidenceof membrane compression to the rootsurface, bone and cementum formationappeared to be reduced. Underminingroot resorption was observed in one site.
Histometric analysis
The histometric analysis is presented inTables 1–7. Table 1 shows healing in all
(a) (b) (c)
Fig. 3. Photomicrographs of supra-alveolar periodontal defect implanted with the occlusive ePTFE membrane. This site experienced wouldfailure and membrane exposure at 9 days postsurgery. The membrane was immediately removed and the mucogingival flaps were adapted andsutured to cover the tissue formed under the membrane. Overview (a; original magnification 2.5� , hematoxylin) and higher magnification (b;4� ) of the lingual aspect of the defect show new bone formation paralleled by regeneration of cementum and a functionally orientedperiodontal ligament (c; 4� , polarized light ). Animal #96.
Periodontal repair in dogs 659
(a) (b) (c)
Fig. 4. Photomicrographs of supra-alveolar periodontal defect implanted with the macroporous ePTFE membrane. Overview (a; originalmagnification 2.5� , hematoxylin) and higher magnification (b; 8� ) of the buccal aspect of the defect show new bone formation paralleledby regeneration of cementum and a functionally oriented periodontal ligament (c; 8� , polarized light). Animal #91.
(a) (b) (c) (e)
(d) (f)
Fig. 5. Photomicrographs of supra-alveolar periodontal defect implanted with the macroporous ePTFE membrane. Overview (a, b; originalmagnification 2.5� , 4� , hematoxylin) and higher magnification (c, d; 8� ) of the buccal aspect of the defect show bone formationparalleled by regeneration of cementum and a functionally oriented periodontal ligament (e, f; 8� , polarized light). Animal #95.
660 Wikesjo et al.
defect sites receiving the occlusiveePTFE membrane. Tables 2 and 3 showhistometric observations in these sitesseparated by wound failure. Membraneexposure and wound failure reduced thepotential for cementum regeneration(Table 4). Table 5 shows healing indefect sites receiving the macroporousePTFE membrane.
Tables 6 and 7 compare observationsin defect sites receiving the occlusiveand macroporous ePTFE membranes.When all defect sites were considered,there were no significant differencesbetween occlusive and macroporousePTFE membranes (Table 6). For theanalysis of the biologic potential forregeneration, however, sites exhibitingmembrane exposure and wound failuremust be excluded from the analysis.Table 7 compares defect sites receivingocclusive and macroporous ePTFEmembranes without wound failure.Bone regeneration approximated 64%of the defect height in sites receiving theocclusive ePTFE membrane and 41% insites receiving the macroporous ePTFEmembrane (p5 0.3113). Cementum re-generation was significantly increasedin sites receiving the occlusive ePTFEmembrane (94% of the defect height)compared with that of the macroporousmembrane (47% of the defect height;p5 0.0167). There were no other sig-nificant differences between the groups.
Discussion
The objective of this study was toevaluate the validity of a concept ofcell occlusion for GTR devices. Macro-porous and occlusive ePTFE mem-branes were surgically implanted incontralateral supra-alveolar periodontaldefects in six Beagle dogs and healingwas allowed to progress for 8 weeksprior to euthanasia and histologic andhistometric evaluation. Three animalsreceiving occlusive ePTFE membranesexperienced wound failure within 2–3weeks postsurgery, resulting in expo-sure and removal of the membranes. Toevaluate the biologic potential of theGTR devices, only animals withoutwound failure and membrane removalwere included. Similar amounts of boneregeneration were observed for animalsreceiving occlusive and macroporousePTFE membranes. Cementum regen-eration was significantly enhanced insites receiving the occlusive ePTFEmembrane compared with sites receiv-ing the macroporous membrane. These
Table 1. Summary statistics for jaw quadrants receiving the occlusive ePTFE membrane
group 5.070.1 5.070.1 2.570.4 2.470.5 0.070.0 0.370.7
in % of the defect height 100 50 48 0 6
Periodontal repair in dogs 661
observations may provide a biologicrationale for tissue occlusion to opti-mize GTR in periodontal defects.
Substantial cementum regenerationwas observed following placement ofthe occlusive and macroporous ePTFEmembranes. Mean cementum regenera-tion approximated 72% and 50% of thedefect height, respectively, exceedingthat observed in previous studies in thisanimal model. After excluding animalsexperiencing membrane exposure, meancementum regeneration in sites receiv-ing the occlusive ePTFE membraneapproached 94% of the defect height.This observation supports the hypoth-esis that tissue-occlusive GTR devicesmay optimize conditions for periodontalregeneration. Nevertheless, observa-tions of substantial cementum regenera-tion including a re-establishedfunctionally oriented periodontal liga-ment in defect sites receiving themacroporous ePTFE membrane suggestthat tissue and/or cell occlusion is not anabsolute requirement for GTR. Sigurds-son et al. (1994, 1995), also using theocclusive ePTFE membrane in the sameanimal model, observed enhanced ce-mentum regeneration including a re-established functionally oriented period-ontal ligament approximating 41% ofthe defect height compared to 1% in thesurgical controls (no membrane) follow-ing an 8-week healing interval. A studyby Wikesjo et al. (2002b) also evaluat-ing the macroporous ePTFE membranein the supra-alveolar periodontal defectmodel showed cementum regenerationapproximating 22% of defect height.Although differences may exist betweenstudies, collectively the studies suggest
that the use of space-providing, occlu-sive or macroporous GTR membranesmay significantly enhance periodontalregeneration.
The newly formed cementum in-cluded functionally oriented fibers irre-spective of whether the defect site hadreceived an occlusive or a macroporousePTFE membrane. This observation issomewhat remarkable since the teethwere sheltered under space-providingePTFE membranes that usually did notcontact the teeth. The membranes, inturn, were submerged under the gingivalflaps, the animals were kept on a softdiet, and the teeth in the opposingmaxillary jaw quadrants had been re-moved or reduced in height to avoid anycontact with the healing defect sites.The functionally oriented periodontalattachment was observed in the absenceand presence of newly formed alveolarbone. Moreover, the newly formedperiodontal attachment had reached thismaturity within an 8-week healinginterval consistent with observationsby Sigurdsson et al. (1994, 1995) andWikesjo et al. (2002b, c). These ob-servations also suggest that the forma-tion and maturation of the periodontalattachment exceeds that of alveolarbone. Intriguingly, a functionally or-iented periodontal attachment was rou-tinely formed without obviousfunctional challenge from occlusion ormasticatory activity.
There was no significant difference incoronal regrowth of alveolar bonebetween defect sites receiving occlusiveor macroporous ePTFE membraneswithout wound failure (3.2 versus2.4 mm, respectively). Bone regenera-
tion in sites with wound failure andearly membrane removal amounted to1.7 mm. Similar observations of boneregeneration were reported by Sigurds-son et al. (1994, 1995). Bone regenera-tion following use of the occlusiveePTFE membrane averaged 2.9 mmfollowing an 8-week healing interval.Mean alveolar bone regrowth in surgicalcontrols without a GTR deviceamounted to 0.6 mm. No bone regen-eration was observed in defects withwound failure and membrane exposure.Wikesjo et al. (2002b) reported a meanregeneration of alveolar bone amount-ing to 2.0 mm in sites receiving themacroporous ePTFE membrane. Collec-tively, the data suggest that space-providing GTR devices, in these studiesexclusively based on a proprietaryePTFE technology, whether occlusiveor macroporous, support substantialregeneration of alveolar bone. Woundfailure, membrane exposure, and infec-tion may severely compromise regen-eration of alveolar bone. However, ifthe GTR device is removed in theimmediate sequence of wound failureand membrane exposure, the regenera-tive potential appears minimally com-promised.
Only one tooth (with an occlusiveePTFE membrane) exhibited ankylosisin this study. Sigurdsson et al. (1994,1995) reported that three of nine andtwo of eight teeth exhibited ankylosis inanimals receiving the occlusive ePTFEmembrane or served as surgical con-trols, respectively. Wikesjo et al.(2002b) reported limited ankylosis intwo of eight teeth in animals receivingthe macroporous ePTFE membrane.Gottlow et al. (1984) did not observeankylosis following GTR in a nonhu-man primate study. Differences inregeneration rates between the tissuessequestered underneath the GTR de-vices, i.e., the alveolar bone and theperiodontal ligament, may account forthese observations. Gottlow et al. (1984)suggested that ‘‘the migration rate ofperiodontal ligament cells is at least ashigh as that of bone cells’’. In fact,tissue resources from the periodontalligament may command an even higherregeneration rate than that of thealveolar bone. Nevertheless, the dataclearly suggest that ankylosis is a rarebut biologic possibility also followingGTR under optimal conditions forwound healing.
Limited root resorption of undermin-ing character was observed in one
Table 6. Comparison between groups (means7SD; mm) receiving the occlusive and themacroporous (P) ePTFE membrane
Table 7. Comparison between groups (means7SD; mm) receiving the occlusive and themacroporous (P) ePTFE membrane excluding animals exhibiting wound failure and earlymembrane removal
animal receiving the macroporousePTFE membrane. Most other teethexhibited minor resorption of surfaceerosion character. There were no appar-ent differences between teeth receivingthe macroporous and occlusive ePTFEmembrane. These observations are con-sistent with those of Sigurdsson et al.(1994, 1995) and Wikesjo et al. (2002a,b, c). In contrast, the teeth in thesurgical controls in the study of Si-gurdsson et al. (1994, 1995) not onlyexhibited limited cementum regenera-tion but also frequent and advanced rootresorption. Collectively, these observa-tions suggest that the migration andproliferation rate of the gingival con-nective tissues are slower than that ofthe periodontal ligament and that onlywhen the gingival connective tissues areallowed to collapse onto and contact theroot surface, more advanced root re-sorption may routinely be observed.
Wound failure and membrane expo-sure has been associated with a de-creased regenerative potential followingGTR. Selvig et al. (1992) showed thatthe extent of oral exposure and bacterialcontamination of an ePTFE membraneat the time of removal may be anindicator of the success or failure ofthe regenerative procedure. Trombelli etal. (1997) showed that intrabony period-ontal defect sites with exposed GTRdevices at the 5-week removal exhibitedminimal, if any, gain of clinical attach-ment. In this study, ePTFE membraneswere exposed and removed as early as9, 11, and 21 days postsurgery. Thedefect sites exhibited substantial regen-eration of cementum and alveolar boneat the 8-week evaluation time-point inspite of the membrane exposure. Im-portantly, the membranes were removedwithin days of exposure. These observa-tions suggest that periodontal regenera-tion has reached sufficient maturity toallow membrane removal within a fewweeks of wound healing. Comparing theobservations in this study with that ofmembrane exposure in other preclinicalstudies suggests that early membraneremoval is critical to maintain anynewly regenerated tissues (Haney et al.1993, Sigurdsson et al. 1994, 1995).
Sites receiving the macroporousePTFE membrane did not experiencewound failure and membrane exposure,indicating that the macroporous struc-ture better supported the healing gingi-val flap than the occlusive ePTFEmembrane. The macroporous mem-brane readily allowed passage of fibro-
vascular elements with limited signs ofinflammation. This tissue ‘‘ingrowth’’may stabilize the wound and secure thenutritional base for the flap during theearly healing sequel, thus preventingwound dehiscence and membrane ex-posure. Similar observations of woundhealing have been made when thismacroporous membrane was used withrhBMP-2 in supra-alveolar periodontal(Wikesjo et al. 2002b) and peri-implant(Wikesjo et al. 2002d) defects. Sixteenanimals received the macroporousmembrane with or without rhBMP-2.None of the defect sites experiencedwound failure and membrane exposureover an 8-week healing interval. In thisstudy, half of the animals receiving theocclusive ePTFE membrane showedwound failure during the early periodof healing, and subsequently exhibitedless bone and cementum regenerationthan those not showing wound failure.These observations are consistent withthose of Sigurdsson et al. (1994, 1995).Two of five animals experienced woundfailure and membrane exposure, how-ever, the membranes were maintained insitu for the 8-week healing interval. Noregeneration of periodontal structureswas observed in these sites. Theseobservations may have general clinicalimplications. A macroporous membranetechnology may be a more predictabledevice for GTR and be more conduciveto uncomplicated clinical managementthan occlusive GTR devices.
Conclusions
� Tissue occlusion does not appear tobe a critical determinant for GTR;substantial bone and cementum re-generation including a functionallyoriented periodontal ligament occursin the presence of space provisionwith or without tissue occlusion.However, tissue occlusion may be arequirement for optimal GTR.
� Periodontal regeneration appears toestablish early (within 9 days).
� Ankylosis and root resorption are rarefindings following GTR.
� Use of macroporous space-providingdevices may increase the predictabil-ity of GTR therapy.
Zusammenfassung
Parodontale Reparation bei Hunden: Gewe-beokklusivitat, eine kritische Bedingung furGTR?Hintergrund: Die Gestaltungskriterien fur
Membranen zur gesteuerten Geweberegenera-tion (GTR) umfassen Biokompatibilitat, Zel-lokklusivitat, Raumerhaltung, Gewebeinteg-ration und Einfachheit in der Handhabung.Vorangegangene Studien haben die Bedeutungder Wundstabilisierung und Raumschaffungwahrend der fruhen Heilungsphase fur erfol-greiche GTR-Therapie herausgestellt wie auchdie Biokompatibilitat, Gewebeintegration, undklinische Anwendbarkeit verschiedener Bioma-terialien untersucht.Zielsetzung: Untersuchung der Bedeutung desGewebeverschlusses als kritischer Einflussfak-tor fur die Resultate nach GTR.Methoden: Supralveolare parodontale Routine-defekte von kritischer Gro�e (5-6 mm) wurdenum die Unterkieferpramolaren von 6 jungenBeaglehunden geschaffen. RaumschaffendeePTFE-Membranen mit (makroporos) oderohne (okklusiv) 300 mm gro�e mit Lasergeschaffenen Poren im Abstand von 0,8 mmwurden als Voraussetzung fur GTR implantiert.Die Membranmodalitaten wurden randomisiertdem rechten oder linken Quadranten bei dennacheinander behandelten Tieren zugewiesen.Die gingivalen Lappen wurden koronalverschoben, um die Membranen zu deckenund vernaht. Die Tiere wurden 8 Wochen nachTherapie fur die histologische und histome-trische Analyse getotet.Ergebnisse: 3 Tiere zeigten einen Wundhei-lungsmisserfolg 2-3 Wochen postoperativ mitExposition und Entfernung der okklusivenePTFE-Membran. Alle Defekte zeigten unab-hangig vom Membrantyp oder der Vor-geschichte einer Exposition und Entfernungdeutliche Hinweise fur parodontale Regenera-tion einschlie�lich funktionell orientierter Des-modontalfasern. Um das biologische Potentialder GTR-Membranen untersuchen zu konnen,wurden nur Tiere ohne Membranexposition und-entfernung in die Analyse einbezogen. Ahn-liche Mengen (Mittelwert7Standardabwei-chung) Regeneration alveolaren Knochenswurden bei Tieren, die okklusive und makro-porose ePTFE-Membranen bekommen hattengefunden (3,271,1 vs. 2,470,5 mm;p5 0,3029). Die Regeneration des Zementswar verbessert in den Defekten, die mitokklusiven ePTFE-Membranen behandelt wor-den waren im Vergleich zu makroporosenMembranen (4,770,4 vs. 2,570,3 mm;p5 0,0186). Ankylose wurde bei einem Tierbeobachtet. Begrenzte Wurzelresorptionen wur-den bei einem weiteren Tier beobachtet.Schlussfolgerung: Gewebeokklusivitat scheintkeine kritische Determinante fur GTR zu sein.Allerdings scheint Gewebeokklusivitat eineBedingung fur optimale Ergebnisse der GTRzu sein. Makroporose platzhaltende Membranenkonnten die Vorhersagbarkeit der GTR-Ther-apie verbessern.
Resume
Reparation parodontale chez les chiens: occlu-sion tissulaire gingivale, necessite critique pourla GTR?Les criteres pour la regeneration tissulaireguidee (GTR) comprennent la biocompatibilite,l’occlusion cellulaire, le maintien de l’espace,
Periodontal repair in dogs 663
l’integration tissulaire et la facilite d’utilisation.Des etudes anterieures ont etabli l’importancede la stabilisation de la plaie et de l’espacedurant les premieres sequences de la guerisonafin de faciliter le processus de GTR et ontevalue la biocomptabilite, l’integration tissu-laire, la maniabilite clinique des differentsbiomateriaux. L’importance de l’occlusion cel-lulaire et tissulaire doit encore etre etablie. Lebut de cette etude a ete d’evaluer le role del’occlusion tissuaire comme determinant cri-tique des resultats de la GTR. Des lesionsparodontales sus-alveolaires de 5-6 mm, ont etecreees autour des premolaires mandibulaires desix jeunes chiens Beagle. Deux types demembranes en teflon ont ete utilisees pour laGTR, avec macro-pores (300mm realises aulaser) ou sans (occlusives). Les traitements ontete repartis dans les quadrants gauche et droitde la mandibule. Les lambeaux gingivaux ontete mis en place pour couvrir les membranes etsutures. Les animaux ont ete euthanasies huitsemaines apres la chirurgie pour l’analysehistologique et histometrique. Trois animauxont eu une mauvaise guerison deux a troissemaines apres la chirurgie vu l’exposition desmembranes en teflon occlusives. Toutes leslesions, quelque soit la configuration ou l’his-torique d’exposition de membranes et d’enleve-ment, ont montre une evidence de regenerationparodontale incluant un ligament parodontalavec des fibres orientees ‘‘fonctionnellement’’.Pour evaluer le potentiel biologique des appa-reils servant a la GTR seuls les animaux sansprobleme de guerison et d’enlevement demembrane ont ete inclus. Des quantites sembl-ables (moyenne7SD) de regeneration osseusealveolaire ont ete observees chez les animauxrecevant des membranes occlusives et macro-poreuses (3,271,1 mm vs 2,4 70,5 mm:p5 0,3029). La regeneration cementaire a eteaugmentee dans les lesions qui ont recu lesmembranes occlusives comparee aux macro-poreuses (4,770,4 mm vs 2,570,3 mm:p5 0,0186). L’ankylose n’a ete observee quechez un animal. Une resorption radiculairelimitee a ete observee chez un autre. L’occlu-sion tissulaire ne semble pas etre un determi-nant critique pour la GTR. Cependant,l’occlusion tissulaire peut etre necessaire pourune GTR optimale. De plus, les systemes degarde d’espace macroporeux peuvent augmen-ter la capacite d’anticipation du traitement GTRclinique.
