The Use of Enamel Matrix Derivative in the Treatment of Periodontal Osseous Defects: A Clinical Decision Tree Based on Biologic Principles of Regeneration

Stuart Froum, DDS*
Jeffrey Lemler, DDS**
Robert Horowitz, DDS***
Bruce Davidson, DDS****

An investigation of the periodontal literature reveals four critical factors necessary for a regenerative response in the treatment of intrabony defects. These include root preparation to remove toxins and altered cementum, space creation by the graft or membrane barrier for migration of progenitor cells, stabilization and flap coverage of the graft, and use of a membrane barrier or enamel matrix derivative (EMD). This article reviews the literature supporting this concept and presents a  clinical decision tree to determine when to use EMD alone or with autogenous grafts and membrane barriers in the treatment of defects of varying morphologies. The clinical decision tree is designed to increase the predictability of a positive clinical response. Cases are shown to demonstrate the indication for each treatment. (Int J Periodontics Restorative Dent 2001;21:437–449.)

****Clinical Professor, Department of Surgical Services (Periodontics) and Department of Implant Dentistry, New York University College of Dentistry.
****Associate Clinical Professor, Department of Surgical Sciences (Periodontics) and Department of Implant Dentistry, New York University College of Dentistry
****Associate Clinical Professor, Department of Implant Dentistry, New York University College of Dentistry
****Associate Clinical Professor, Department of Surgical Sciences (Periodontics), New York University College of Dentistry
****Reprint requests: Dr Stuart Froum, 17 West 54th Street, Suite 1 C/D, New York, New York 10019.

Since Melcher1 first advanced the biologic principles associated with guided tissue regeneration, clinicians and researchers have sought techniques to predictably regenerate periodontal structures lost to the disease process. Successful regeneration, which can only be evidenced histologically, consists of a functionally oriented periodontal ligament (PDL) inserting into newly formed cementum and alveolar bone. A review of the periodontal literature suggests that surgical procedures using autografts, allografts, xenografts, and/or barrier membranes can contribute to a successful regenerative response.2–5 Because of the obvious difficulties in obtaining human post treatment histology, other measurement parameters have become the standard for evaluating a healing response. From a clinical viewpoint, measurements such as gain in probing attachment level, decrease in probing depth, fill of the osseous defect, and percentage fill of the osseous defect have been used to measure and compare the results of various “regenerative” therapies. In cases where the clinical response is positive and histologic evidence (human block sections) shows that regeneration is possible, the term “regenerative therapy” may then be applied to that specific material or technique.

Analysis of published regenerative studies shows that four critical factors are consistent requirements. First, the root surface must be prepared so that bacterial cells and other cytotoxic products are removed. 6–9 Second, space must be created so that progenitor cells from the PDL can selectively repopulate the root surface.10–12 Third, stability of the maturing fibrin clot, graft, implant, and/or membrane under the tissue flap is essential. 13–15 Fourth, the flap must be designed to maintain coverage of the barrier membrane and/or graft for containment of the material to prevent bacterial contamination and to minimize recession (Fig 1).16–20

Enamel matrix derivative (EMD) (Emdogain, Biora) was introduced into the periodontal literature in 1997 as a tissue-healing modulator to mimic events that occur during root development and to help stimulate periodontal regeneration.21–23 The use of EMD in the treatment of intrabony defects has been shown to result in significant improvement in probing depth, clinical attachment level (CAL), and bone fill.24–29 Histologic evidence of regeneration of new periodontal attachment
has been demonstrated in human and animal model systems.30–35 Two separate studies compared the benefits of EMD to those obtained with the use of bioabsorbable and nonabsorbable barrier membranes. 27,33 The results showed no significant difference in probing depth reduction, CAL gain, and gingival recession between the membrane- and EMD-treated sites. Furthermore, all test groups (EMD and membrane treatment) showed a superior result when compared to the open-flap debridement control. At the time of this writing, there were no data on the use of EMD in conjunction with barrier membranes and/or bone grafts and only one published study on the use of EMD in combination with a xenograft.36

The recommended surgical technique and postoperative care for using the EMD material have been previously described.29,32 In our experience, and in a review of the periodontal literature of EMD treated cases, the recommended technique has demonstrated variable clinical and histologic outcomes. 34 Clinical results are often dependent upon the dimension and morphology of the intrabony defect. Deeper intraosseous lesions result in greater gains in CAL and bone fill than shallower defects. Pontoriero et al27 and Heden et al26 have documented this difference in the healing response of intrabony defects treated with EMD. Previous studies with other regenerative techniques showed variable healing responses dependent upon the number of walls in an osseous defect. 37–39 Three-walled defects have a greater potential to fill than one- and two-walled defects. Others have proposed that the potential for fill of an osseous defect is directly related to the number of surrounding osseous walls, the amount of exposed root surface, and the ability to obtain flap coverage of the treated area.40 Following surgery, it has also been shown that there is a greater gain of attachment when the angle of the defect to the long axis of the root is less than 45 degrees.41

Figure 1            “4S pyramid: of regenerative requirements
Figure 2            Clinical decision tree for treatment of periodontal defects with EMD

Clinical decision tree and criteria for treatment section
The basis for successful use of the EMD material has been the precipitation of the protein along the root surface approximating the osseous defect and the resulting stimulation of new cementum. It is therefore logical to assume that a limiting factor for success would include the ability to contain the EMD adjacent to the affected root surface. A clinical decision tree (CDT) (Fig 2) combines the biologic principles of regeneration and considerations of defect morphology in an effort to improve clinical success. Modifications of the surgical approach are presented to facilitate the use of EMD for defects of varying morphologies.

There are three branches on the EMD decision tree. Each allows the clinician to evaluate the depth and morphology of a specific defect and plan the appropriate treatment sequence. The first branch involves well-contained defects, ie, two- and three-walled intrabony defects and craters. The morphology of these defects permits the EMD to be well contained against the affected root surface.

Case 1 showed a 6-mm-deep two-walled defect on the distal aspect of the mandibular right first molar. The patient was a 24-year-old nonsmoking man with a negative medical history. The flap was reflected, and the soft tissue defect was debrided. At the time of photographing the initial osseous defect, it was noted that blood had pooled within the walls. A photograph was taken of the level at which the defect was able to contain the blood (Fig 3a). The root was planed, the defect irrigated and dried, and the smear layer removed with citric acid applied for 15 seconds. The root was then rinsed with sterile saline for 1 minute and again dried. The EMD was syringed along the root surface, starting from the base of the defect and extending coronally. At the time of flap closure, the EMD was well contained within the entire 5.5-mm intraosseous defect. This was similar to the level of the pooled blood in the initial osseous defect as seen in Fig 3a. Twelve months postsurgery, the surgical site was reentered and the residual depth of the defect was recorded. Bone had filled to a level equal to the most coronal position of the EMD at the time of closure. Fill of the defect measured 3.4 mm (approximately 62% of the original defect) (Fig 3b). Presurgical and postsurgical radiographs recorded the defect fill (Fig 3c).

The second branch of the CDT involves deep, wide intraosseous defects. The morphology of these defects would permit only limited containment of the EMD. With these types of lesions, an autogenous bone graft is advised in an effort to hold the EMD material against the affected root surface and help maintain space by supporting the flap. Following placement of the EMD, an autogenous graft obtained from intraoral edentulous areas (or ramus) is placed into the defect. No attempt is made to dry or condense the graft material. The surgical flaps are repositioned or positioned coronally in an attempt to achieve primary closure.

Figure 3a           Case 1. A 6-mm two-walled defect on the distal aspect of the mandibular right first molar after debridement and prior to EMD treatment. Note the level of the pooled blood.
Figure 3b           Same site at 12-month reentry. The defect filled to the level that the EMD filled the defect, the same level as the pooled blood.
Figure 3c           Presurgical radiograph of the distal aspect of the first molar (left) and 1-year postsurgical radiograph prior to reentry surgery (right).

Case 2 illustrates treatment of this type of defect. The patient was a healthy, nonsmoking 46-year-old woman who presented with clinical and radiographic evidence of severe, chronic periodontitis. Initial therapy consisted of a review and reinforcement of personal plaque control procedures, root planning, and selective occlusal equilibration. Eight weeks following the initial treatment, a 10-mm defect was measured on the mesiopalatal aspect of the maxillary right canine.
Eight mm of probing depth was measured on the distopalatal aspect of the maxillary right lateral incisor. Upon debridement of the osseous defects, intrabony depths were measured at 6 mm on the mesial aspect of the canine and 5 mm on the distal aspect of the lateral incisor (Fig 4a). The anatomy of the defects would not permit full containment of the EMD against the root surface.
An autogenous graft was taken from the ridge and maxillary tuberosity distal to the right second molar. After the application of EMD, the graft was placed. The EMD was then reapplied over the bone graft (an EMD “sandwich”) (Fig 4b). One year following the grafting procedure, the probing depths were 3 mm. A reentry procedure showed that the 5 mm of osseous fill on the mesial aspect of the canine and 4 mm on the distal aspect of the lateral incisor had eliminated the intrabony defect (Fig 4c). Presurgical and postsurgical radiographs reflected these changes (Fig 4d).

The final two cases illustrate the third branch of the decision tree and involve treatment of horizontal bone loss or shallow infrabony defects. Similar to the treatment of lesions associated with the other two branches of the decision tree, full thickness mucoperiosteal flaps were elevated, the osseous defects debrided, and the root surfaces planed. The smear layer in one case was removed with 24% ethylenediaminetetraacetic acid (EDTA) applied for 2 minutes and in the other case citric acid pH 1 applied for 15 seconds. The roots were then irrigated and dried, and EMD  as applied. The autogenous graft material was placed, and EMD was again applied over the bone graft. For these horizontal or shallow defects, the graft material is covered with a bioabsorbable membrane to further contain the material. The membrane was stabilized with bone tacks.

Figure 4a           (left) Case 2. The debrided osseous defects measure 6 mm on the mesiolingual aspect of the maxillary right canine and 5 mm on the distolingual aspect of the lateral incisor.
Figure 4b           (right) Defects are filled with autogenous bone and covered with EMD.
Figure 4c           (left) One-year reentry reveals a 5-mm fill on the mesiopalatal aspect of the canine and 4 mm of fill on the distopalatal aspect of the lateral incisor.
Figure 4d           (right) Presurgical radiograph of the canine and lateral incisor (left) and the 1-year postsurgical radiograph (right).

Case 3 involved the maxillary right canine of a healthy, nonsmoking 65-year-old woman who had a negative medical history. Thecanine served as the terminal abutment  for an anterior fixed partial denture, which supported a bilateral, distal extension, semiprecision, removable partial denture. The trough-like osseous defect extendedfrom the distobuccal to the  distal of the palatal aspect of the tooth, creating a shallow intrabony defect with an extensive Supracrestal component (Fig 5a). Following placement of the EMD material along the root surface adjacent to the osseous defect, an autogenous graft was placed into the site. The graft was obtained with a trephine from the maxillary right edentulous ridge, milled, and carefully placed to fill the defect. EMD was then placed on the graft surface. The graft was covered with a bioabsorbable membrane (BioGide, Osteohealth) and stabilized on the buccal and palatal aspects with nonresorbable bone tacks (Ace Surgical Supply). The tissue flaps were coronally positioned to obtain primary closure over the membrane. At 1 year, the probing depth had been reduced from 11 to 4 mm. The 7-mm osseous defect had completely filled. There was 3 mm of supracrestal bone apposition (Fig 5b). Presurgical and postsurgical radiographs demonstrated the osseous changes (Fig 5c).

Case 4 involved buccal and mesial defects adjacent to the maxillary right central incisor of a healthy, nonsmoking 50-year-old man with no significant medical history. Probing depths measured 13 mm on the buccal and 14 mm on the mesial aspect of the maxillary right central incisor (Fig 6a). The osseous defect was found to be 10 mm on the buccal and 11 mm on the mesial aspects (Fig 6b). The tooth also evidenced Miller Class III mobility. Prior to undertaking the surgery, the tooth was stabilized with a composite splint joining it to the adjacent lateral incisor. The source of the autogenous intraoral graft was bone removed during an osteotomy for the placement of a dental implant in the left central incisor position. Additional bone was obtained from the right tuberosity. A bioabsorbable membrane (BioGide) was placed over the bone graft and stabilized with bone tacks (Fig 6c). The flap was coronally positioned to submerge the membrane. At the 6-month postoperative evaluation, probing depth had been reduced
to 3 mm. The reentry photograph showed 5 and 6 mm of Supracrestal bone growth on the buccal and mesial aspects, respectively (Fig 6d).

Figure 5a           Case 3. The distal osseous defect on the maxillary right canine following debridement demonstrates an intrabony and suprabony defect. The defect was filled with EMD, autogenous intraoral bone, and a bioabsorbable membrane barrier. The barrier was stabilized with bone tacks.
Figure 5b           One-year reentry reveals the 7-mm defect filled completely and 3 mm of supracrestal bone apposition.
Figure 5c           Presurgical radiograph of the distal defect on the maxillary right canine (left) and 1-year postsurgical radiograph (right).

Discussion
Each of the cases presented illustrates the application of the CDT for treatment of periodontal defects with EMD. While this decision tree is based on defect size and morphology, it also adheres to the critical requirements for attaining successful clinical results. These factors—surface treatment, space, stability, and submerged material (flap coverage)— have been documented in the literature as contributing factors to a positive healing response.

Root surface preparation is necessary to minimize the presence of bacteria and endotoxins. Contamination of the root surface of periodontally compromised teeth has been well documented in the literature. Scaling and root planning have been shown to remove the endotoxins contaminating a periodontally involved root surface.42 The use of root-surface modifiers remains controversial. Citric acid43 and an etching agent at neutral pH (24% EDTA)44 have demonstrated successful root demineralization and the creation of a root surface that favors fibroblastic attachment. However, a meta-analysis of guided tissue regeneration studies showed no clinical advantage with the use of citric acid conditioning.45 With EDTA, there is 10% less recession, 10% greater reduction of pocket depth, and more total histologic attachment when compared to citric  acid and a control.46 The use of root conditioning with EMD follows the manufacturer’s recommendation. This protocol remains to be tested to determine if and when conditioning is a necessary requirement following root planning and prior to the EMD application.

A primate study47 and a dog model48 with Class III furcation defects demonstrated the importance of space maintenance and showed that little or no regeneration occurs if the membranes collapse
into the defect sites. Articles reviewing evidence-based approaches to the successful regeneration of intrabony defects18 and mandibular Class II furcation defects49 and a study on Class III furcation repair in dogs50 cited the importance of space maintenance. Regenerative studies12 concluded that the greatest gains of “newly formed granulation tissue” are found in deeper defects because there is greater space that could be filled with a blood clot.

Figure 6a           Case 4. Buccal probing of maxillary right central incisor reveals an 11-mm pocket depth.    
Figure 6b           (right) Exposed and debrided defect measures 10 mm on the buccal and 11 mm on the mesial aspects.
Figure 6c           Autogenous bone obtained during the osteotomy to prepare the left central incisor implant site is placed in the defect on the right central incisor after treatment with EMD and covered with a bioabsorbable membrane barrier.
Figure 6d           (right) Six-month reentry reveals 5 and 6 mm of supracrestal fill on the buccal and mesial aspects, respectively.

Stability of the clot and bone graft has been shown to be a basic requirement in all wound-healing studies.51–53 In the periodontal literature, stability of the clot, membrane, and tooth have been discussed as components of a successful healing response. Wikesjö et al14 reviewed the significant events in the early healing of a periodontal wound and stressed the importance of a stable clot and wound if regeneration is to succeed. The 1996 World Workshop in Periodontics54–56 reiterated that mobility of a wound margin may be a potential cause of disruption of the fibrin clot and a reason for procedure failure. Dog models demonstrated that root surfaces conditioned with heparin lack wound stability, resulting in minimal connective tissue repair along the root surface.15,48 An 8-year longitudinal study57 found that mobile teeth do not respond as well as nonmobile teeth to periodontal treatment. A 2- year study58 reported that “occlusal adjustment in conjunction with conventional periodontal therapy resulted in a more favorable clinical attachment level for the adjusted than the non-adjusted patients.”

Lastly, flap coverage of the membrane barrier has been cited as essential for optimum healing. Murphy59 stated that 85% of all membranes used in his study became exposed and that exposure time was approximately 2 weeks after placement. Membrane exposure has been associated with a reduction in attachment gain.60,61 The greater the exposure of the membrane, the less the gain of clinical attachment. It has been demonstrated that tissue flaps with a thickness of 1 mm or less are associated with 3.5 times more flap recession than those with flaps thicker than 1 mm.62 Coronal flap repositioning has been shown to increase the chances of supracrestal apposition of new bone.63

The CDT presented in this article is designed to increase the predictability of obtaining a positive clinical result when using EMD to treat osseous defects with varying morphologies. EMD, without the addition of bone substitutes and/or membrane barriers, has been demonstrated to be clinically successful in reducing probing depth, gaining CAL, and volumetrically increasing the defect fill in the treatment of intraosseous defects (Table 1). However, analysis of the results of clinical studies using EMD alone demonstrates the indications of this approach.

Heijl et al24 evaluated the treatment of one- or two-walled osseous defects. They noted that the “mean intrabony component” was “close to 5 mm, with about one quarter of the defect having a 3 wall component.” Apparently, approximately half of the reported fill occurred in the well-contained part of the defect. This is in accordance with the first branch of the CDT.

A study of 145 defects treated with EMD reported a CAL gain of 4.6 mm and probing depth reduction of 5.2 mm.26 The recorded fill was 2.9 mm on average, which represented 69% of the original defect. Sixty-six percent of the treated defects were two walled, two and three walled, or three walled. Although these types of defects offer excellent containment of the EMD gel and respond well to EMD alone (as indicated by the CDT), the authors concluded that neither the type nor depth of the intrabony defect reached statistical significance when correlating them with the clinical results.

Another study of 72 consecutively treated intrabony defects of one- or two-walled morphology28 reported a mean pocket depth reduction of 4.7 mm and a mean CAL gain of 4.2 mm at 12 months postsurgery. It also reported a mean radiographic fill of 3.1 mm and mean defect fill of 70%. Although at 1 year postsurgery only smoking and bleeding on probing significantly correlated with defect fill, 69% of the treated defects were two walled and showed a slightly greater bone gain than the one-walled defects (3.3 vs 2.6 mm).

In a 12-month clinical study, 27 osseous defects that were treated with three different membrane barriers or EMD alone were compared to flap debridement controls. The authors noted no significant difference between the membrane and EMD groups, while there was a statistically significant superiority of both of these groups compared to the flap debridement control. The authors further noted that “clinical improvements were better at sites with deep rather than at sites with  shallow intrabony defects.” This held true for all test and control groups and is in accordance with the indications for use of EMD alone as described by the CDT.

MB= membrane barrier; OD= open debridement; EMP= enamel matrix derivative; BPBM= bovine porous bone mineral

A meta-analysis was conducted to evaluate studies that treated intrabony defects with open-flap debridement with and without autogenous and allogeneic bone grafts, and with guided tissue regeneration with barrier membranes.64 Regardless of the treatment modality, bone fill correlated significantly with the depth of the defect. Deeper initial lesions were associated with clinically and statistically improved results.64 The correlation that the deeper the initial intraosseous defect, the more improved the healing result (ie, gain in CAL and defect fill) has been shown to be true with a variety of treatment techniques. Studies using citric acid root conditioning and replaced-flap surgery65 with nonresorbable membrane barriers,66 with bioabsorbable membrane barriers, 67 and with a combination of root conditioning, bone graft, and membrane barriers68 have all  demonstrated a positive relationship between defect fill and initial osseous defect depth. An evidence based literature review of factors that influence clinical and histologic regeneration also concluded that defect morphology can directly affect the predictability of regeneration. The authors noted that deep and narrow defects are the most predictable in achieving a positive regenerative response.49 Considering that almost all defects, regardless of predominant classification, are combination defects with three walls at the base, this correlation may reflect the predisposition for fill of the three-walled component of the osseous defect.35 The fill rate reflects the defect proximity to PDL precursor cells and the osteoconductive nature of the three surrounding osseous walls. In addition to the factors mentioned, the ability of the three-walled component of the defect to contain the EMD against the root surface should promote greater percentage of fill.

Although EMD has been shown in numerous reports to improve clinical parameters and to mediate a regenerative healing response, the mechanism by which EMD exerts its influence on cells has only recently been evaluated. An in vitro investigation of the effects of EMD on the behavior of human PDL cells (HPLF) and gingival fibroblasts (HGF) indicated that both cell types release significantly higher levels of transforming growth factor (TGF) under the influence of EMD than in control cultures. However, whereas HGF was shown to attach poorly and not spread on EMD-coated substrata, HPLF attaches and spreads within 24 hours. The authors concluded that this difference might account for the ability of EMD to promote selective cell repopulation during early stages of periodontal healing.69 Several in vitro studies have reported the effect of EMD on osteoblast precursors and differentiated osteoblasts. One such study indicated that EMD increases proliferation of these two cell types.70 However, while EMD stimulates activity of preexisting osteoblasts, it does not stimulate differentiation of the precursor cells. In a second study, EMD was found to stimulate release of TGF-B1 and increase the expression of a differentiated osteoblastic phenotype.71 These results can be related to another study that showed that cultured PDL cells stimulated by EMD provoke a strong and lasting interleukin- 6 expression that precedes TGF-B1 release.72 Thus, EMD may provoke signaling molecules, which result in TGF-B1 expression 72 hours after exposure to EMD.72 Another in vitro study demonstrated that “PDL cell wound fill rates” are increased significantly compared to those of gingival fibroblasts when EMD is added to a medium containing both cell types. The authors concluded that enhancement of periodontal regeneration may be a result of modifying PDL cell proliferation and migration.73

Two other in vitro studies documented the effects of EMD on osteoblastic cells. In the first, EMD was cultured with bovine PDL cells and rat bone marrow cells. In these cultures, EMD stimulated mineralized nodule formation in a dose dependent manner. This corresponded to an increase in alkaline phosphatase activity.74 In the second study, EMD enhanced the proliferation and differentiation of human osteoblasts.75

These documented effects of EMD on PDL and osteoblast cells, as well as on TGF-_1 release and alkaline phosphatase, may explain the enhanced osteogenesis and PDL regeneration seen in human periodontal defects. From a clinical standpoint, this molecular activity requires space for these cells to proliferate and mediate regeneration. If the defect is not naturally space maintaining, a bone graft should be used. This is described by the second branch of the CDT. The bone graft prevents the flap from collapsing and maintains the EMD in the defect site. This allows the previously described molecular reactions to take place. Although the cases shown here representing the second branch of the CDT used only autogenous bone as the grafting material, clinical and histologic trials are necessary to determine whether allografts and/or  bone replacement grafts (xenografts and alloplasts) in conjunction with EMD may result in similar clinical results.

A recent 6-month reentry study36 comparing 42 bilateral defects—21 treated with EMD alone and 21 with EMD and bovine derived anorganic bone (xenograft)— reported a significant  advantage in probing depth reduction,  CAL gain, and defect fill in the EMD xenograft– treated defects. Defect fill averaged 30% in the defects treated with EMD only, versus 81% in the combination EMD xenograft– treated sites. The majority of the defects in both groups were two walled, which would explain the results. These are consistent with the second branch of the CDT.36

Finally, in shallow, horizontal, or furcation defects, an additional benefit may be derived by placing  a membrane barrier over the EMD– bone graft combination. This combination (third branch of the CDT) requires coronal flap positioning to cover the membrane. The efficacy of coronal flap positioning over a membrane barrier was demonstrated in a human histology study.63 The authors showed supracrestal apposition of new bone, ligament, and cementum. A recent study of surgically created Class II furcation defects in mongrel dogs showed no advantage to the use of EMD in conjunction with modified Widman flap surgery.76 According to the CDT, the treatment that was used followed the first branch (EMD alone). However, the defects being treated required a combination approach, as noted in the second or third branch of the CDT. Further investigation is needed to determine the response of EMD in the treatment of furcation defects with or without the use of bone grafts and membrane barriers.

Adherence to the four critical factors discussed and the therapeutic approach described in the CDT should allow the clinician to observe improved results with greater predictability. Although the CDT approach uses the critical factors involved in the regeneration of a variety of osseous defects, controlled clinical data are still necessary to test the CDT hypothesis.

References
1. Melcher AH. On the potential repair of periodontal tissues. J Periodontol 1976;16:545–548.
2. Garrett S. Periodontal regeneration around natural teeth. Ann Periodontol
1996;1:621–666.
3. Froum S, Gomez C. Periodontal regeneration. Curr Opin Periodontol 1993;1:
111–128.
4. Reynolds MA, Bowers GM. Periodontal regeneration following surgical therapy.
Curr Opin Periodontol 1996;3:126–139.
5. Mellonig JT. Human histologic evaluation of a bovine-derived bone xenograft
in the treatment of periodontal osseous defects. Int J Periodontics Restorative
Dent 2000;20:19–29.
6. Adriaens PA, Edwards CA, De Boever JA, Loesche WJ. Ultrastructural observations on bacterial invasion in cementum and radicular dentin of periodontally diseased human teeth. J Periodontol 1988;59: 493–503.
7. Aleo JJ, De Renzis FA, Farber PA, Varboncoeur AP. The presence and  biologic activity of cementum-bound endotoxin. J Periodontol 1974;45:672–675.
8. O’Leary T. The impact of research on scaling and root planing. J Periodontol 1986; 57:69–75.
9. Nishimine D, O’Leary TJ. Hand instrumentation versus ultrasonics in the
removal of endotoxins from root surfaces. J Periodontol 1979;50:345–349.
10. Machtei EE, Cho MI, Dunford R, Norderyd J, Zambon JJ, Genco RJ. Clinical, microbiological, and histological factors which influence the success of regenerative periodontal therapy. J Periodontol 1994;65: 154–161.
11. Tonetti MS, Pini Prato GP, Cortellini P. Factors affecting the healing response of intrabony defects following guided tissue regeneration and access flap surgery. J Clin Periodontol 1996;23:548–556.
12. Cortellini P, Pini Prato G, Tonetti MS. Periodontal regeneration of human infrabony defects. II. Re-entry procedures and bone measures. J Periodontol 1993;64: 261–268.
13. Wikesjö UME, Selvig KA. Periodontal wound healing and regeneration.
Periodontol 2000 1999;19:21–39.
14. Wikesjö UME, Nilvéus RE, Selvig KA. Significance of early healing events on
periodontal repair. A review. J Periodontol 1992;63:158–165.
15. Wikesjö UME, Claffey N, Egelberg J. Periodontal repair in dogs: Effect of
heparin treatment of the root surface. J Clin Periodontol 1991;18:60–64.
16. Nowzari H, MacDonald ES, Flynn J, London RM, Jorrison J, Slots J. The
dynamics of colonization of barrier membranes for guided tissue regeneration. J
Periodontol 1996;67:694–702.
17. Mombelli A, Lang NP, Nyman S. Isolation of periodontal species after guided tissue regeneration. J Periodontol 1993;64: 1171–1175.
18. Machtei EE, Dunford R, Grossi SG, Genco RJ. Gingival recession and exposure of barrier membrane: Effect on guided tissue regeneration of Class II furcation defects. Int J Periodontics Restorative Dent 1995;15:590–599.
19. Lindhe J, Pontoriero R, Berglundh T, Araujo M. The effect of flap management and bioresorbable occlusive devices in GTR treatment of degree III furcation defects. An experimental study in dogs. J Clin Periodontol 1995;22:276–283.
20. Pontoriero R, Nyman S, Ericsson I, Lindhe J. Guided tissue regeneration in surgically- produced furcation defects. An experimental study in the beagle dog. J Clin Periodontol 1992;19:159–163.
21. Hammarström L. Enamel matrix, cementum development and regeneration. J
Clin Periodontol 1997;24:658–668.
22. Gestrelius S, Andersson C, Lidström D, Hammarström L, Somerman M. In-vitro studies on periodontal ligament cells and enamel matrix derivative. J Clin
Periodontol 1997;24:685–692.
23. Zetterström O, Andersson C, Eriksson L, et al. Clinical safety of enamel matrix derivative (Emdogain) in the treatment of periodontal defects. J Clin Periodontol 1997;24:697–704.
24. Heijl L, Heden G, Svardström G, Östgren A. Enamel matrix derivative (Emdogain) in the treatment of intrabony periodontal defects. J Clin Periodontol 1997;24: 705–714.
25. Sculean A, Reich E, Chiantella GC, Brecx M, Odont L. Treatment of intrabony periodontal defects with an enamel matrix protein derivative (Emdogain): A report of 32 cases. Int J Periodontics Restorative Dent 1999;19:157–163.
26. Heden G, Wennström J, Lindhe J. Periodontal tissue alterations following
Emdogain treatment of periodontal sites with angular bone defects. A series of
case reports. J Clin Periodontol 1999;26: 855–860.
27. Pontoriero R, Wennström J, Lindhe J. The use of barrier membranes and enamel matrix proteins in the treatment of angular bone defects. A prospective controlled clinical study. J Clin Periodontol 1999;26:833–840.
28. Heden G. A case report study of 72 consecutive Emdogain-treated intrabony
periodontal defects: Clinical and radiographic findings after 1 year. Int J Periodontics Restorative Dent 2000;20: 127–139.
29. Froum S, Weinberg MA, Tarnow D. A comparison study utilizing open flap
debridement with and without enamel matrix derivative in the treatment of  periodontal intrabony defects—A 12 month re-entry study. J Periodontol 2001;72:
25–34.
30. Heijl L. Periodontal regeneration with enamel matrix derivative in one human
experimental defect. A case report. J Clin Periodontol 1997;24:693–696.
31. Hammarström L, Heijl L, Gestrelius S. Periodontal regeneration in buccal dehiscence model in monkeys after application of enamel matrix proteins. J Clin
Periodontol 1997;24:669–677.
32. Mellonig J. Enamel matrix derivative for periodontal reconstructive surgery:
Technique and clinical and histologic case report. Int J Periodontics Restorative Dent 1999;19:9–19.
33. Sculean A, Donos N, Windisch P, et al. Healing of human intrabony defects following treatment with enamel matrix proteins or guided tissue regeneration. J
Periodontal Res 1999;34:310–322.
34. Yukna R, Mellonig J. Histologic evaluation of periodontal healing in humans following regenerative therapy with enamel matrix derivative. A 10-case series. J
Periodontol 2000;71:752–759.
35. Sculean A, Chiantella GC, Windisch P, Donos N. Clinical and histologic evaluation of human intrabony defects treated with an enamel matrix protein derivative (Emdogain). Int J Periodontics Restorative Dent 2000;20:375–381.
36. Lekovic V, Camargo PM, Weinlaender M, Nedic M, Aleksic Z, Kenney EB. A comparison between enamel matrix proteins used alone or in combination with bovine porous bone mineral in the treatment of intrabony periodontal defects in humans. J Periodontol 2000;71:1110–1116.
37. Pritchard J. The intrabony technique as a predictable procedure. J Periodontol 1957;28:202–216.
38. Pritchard J. The diagnosis and management of vertical bony defects. J
Periodontol 1983;54:29–35.
39. Ellegaard B, Löe H. New attachment of periodontal tissues after treatment of
intrabony lesions. J Periodontol 1971; 42:648–652.
40. Hiatt WH, Schallhorn RG. Intraoral transplants of cancellous bone and marrow in periodontal lesions. J Periodontol 1973; 44:194–208.
41. Steffenson B, Weber H-P. Relationship between the radiographic periodontal
defect angle and healing after treatment. J Periodontol 1989;60:248–254.
42. Aleo J, De Renzis F, Farber P. In vitro attachment of human gingival fibroblasts to root surfaces. J Periodontol 1975;46: 639–645.
43. Register AA, Burdick FA. Accelerated reattachment with cementogenesis to
dentin, demineralized in situ. I. Optimum range. J Periodontol 1975;46:646–655.
44. Blomlof J, Blomlof I, Lindskog S. Effect of different concentrations of EDTA on smear removal and collagen exposure in periodontitis-affected root surfaces. Int J Periodontics Restorative Dent 1997;17: 242–249.
45. Machtei E, Schallhorn R. Successful regeneration of mandibular Class II furcation defects: An evidence-based treatment approach. Int J Periodontics
Restorative Dent 1995;15:146–167.
46. Blomlof J, Jansson L, Blomlof L, Lindskog S. Root surface etching at neutral pH promotes periodontal healing. J Clin Periodontol 1996;23:50–55.
47. Caton J, Wagener C, Polson A, et al. Guided tissue regeneration in interproximal defects in the monkey. Int J Periodontics Restorative Dent 1992;12:
266–277.
48. Haney J, Nilvéus R, McMillan P, Wikesjö UME. Periodontal repair in dogs: Expanded polytetrafluoroethylene barrier membranes support wound stabilization
and enhance bone regeneration. J Periodontol 1993;64:883–890.
49. Cortellini P, Bowers GM. Periodontal regeneration of intrabony defects: An evidence- based treatment approach. Int J Periodontics Restorative Dent 1995;15: 128–145.
50. Araujo MG, Berglundh T, Lindhe J. GTR treatment of degree III furcation defects with 2 different resorbable barriers. An experimental study in dogs. J Clin
Periodontol 1998;25:253–259.
51. Hulth A. Current concepts of fracture healing. Clin Orthop 1989;249:265–284.
52. Irvin TT. Wound Healing, Principles and Practice. London: Chapman and Hall,
1981.
53. Cohen I, Diegelmann R, Lindblad W. Wound Healing: Biochemical and Clinical Aspects. Philadelphia: WB Saunders, 1992:362.
54. American Academy of Periodontology. Ann Periodontol 1996;1:629–634.
55. Egelberg J. Regeneration and repair of periodontal tissues. J Periodontal Res
1987;22:233–242.
56. Garrett S, Bogle N. Periodontal regeneration: A review of flap management.
Periodontol 2000 1993;1:100–108.
57. Fleszar TJ, Knowles JW, Morrison EC, Burgett FG, Nissle RR, Ramfjörd SP. Tooth mobility and periodontal therapy. J Clin Periodontol 1980;7:495–505.
58. Burgett FG, Ramfjörd SP, Nissle RR, Morrison EC, Charbeneau TD, Caffesse RG. A randomized trial of occlusal adjustment in the treatment of periodontitis patients. J Clin Periodontol 1992;19: 381–387.
59. Murphy KG. Postoperative healing complications associated with Gore-Tex periodontal material. Part II. Effect of complications on regeneration. Int J Periodontics Restorative Dent 1995;15:548–561.
60. Nowzari H, Slots J. Microorganisms in polytetrafluoroethylene barrier membranes for guided tissue regeneration. J Clin Periodontol 1994;21:203–210.
61. Nowzari H, Matian F, Slots J. Periodontal pathogens of PTFE membrane for GTR inhibit healing. J Clin Periodontol 1995; 22:469–474.
62. Anderegg CR, Metzler DG, Nicoll BK. Gingiva thickness in guided tissue regeneration and associated recession at facial furcation defects. J Periodontol 1995;66: 397–402.
63. Stahl SS, Froum S. Human Suprabony healing responses following root demineralization and coronal flap anchorage. Histologic responses in 7 sites. J
Clin Periodontol 1991;18:685–689.
64. Laurell L, Gottlow J, Zybutz M, Persson R. Treatment of intrabony defects by different surgical procedures. A literature review. J Periodontol 1998;69:303–313.
65. Renvert S, Garrett S, Nilvéus R, Chamberlain AD, Egelberg J. Healing after
treatment of periodontal intraosseous defects. VI. Factors influencing the healing
response. J Clin Periodontol 1985;12: 707–715.
66. Tonetti MS, Pini-Prato G, Cortellini P. Periodontal regeneration of human intrabony defects. IV. Determinants of healing response. J Periodontol 1993;64:
934–940.
67. Falk H, Laurell L, Ravald N, Teiwik A, Persson R. Guided tissue regeneration
therapy of 203 consecutively treated intrabony defects using a bioabsorbable
matrix barrier. Clinical and radiographic findings. J Periodontol 1997;68:571–581.
68. Selvig KA, Kersten BG, Wikesjö UME. Surgical treatment of intrabony periodontal defects using expanded polytetrafluoroethylene barrier membranes:
Influence of defect configuration on healing response. J Clin Periodontol 1993;64: 730–733.
69. Van der Pauw MT, Van den Bos T, Everts V, Beertsen W. Enamel matrix-derived protein stimulates attachment of periodontal ligament fibroblasts and
enhances alkaline phosphatase activity and transforming growth factor _1 release of periodontol ligament and gingival fibroblasts. J Periodontol 2000;71:31–43.
70. Pulliam RP, Carnes AAL, Schwartz Z, Mellonig JT, Boyan BD, Cochran DL.
Bone cell response to enamel matrix derivative in vitro [abstr 2988]. J Dent Res
2000;79:517. 
71. Schwartz Z, Cochran DL, Lohmann CH, Carnes DL, Dean DD, Bryan BD.
Osteoblast phenotypic expression is moderated by enamel matrix protein
[abstr 2987]. J Dent Res 2000;79:517.
72. Lyngstadaas SP, Andersson C, Lundberrg E, Ekdakal H. Interleukin expression in cultured human PDL cells growing on Emdogain [abstr 2986]. J Dent Res 2000;79:517.
73. Hoang AM, Oates AW, Cochran DL. In vitro wound healing response to enamel matrix derivative [abstr 1354]. J Dent Res 2000;79:313.
74. Yoneda S, Kasugal S, Itoh D, et al. Effect of enamel matrix derivative (Emdogain) on osteoblastic cell [abstr 2641]. J Dent Res 2000;79:474.
75. Satomura K, Hayashi E, Yamanouchi K, et al. Effects of Emdogain on proliferation and differentiation of human osteoblasts [abstr 2642]. J Dent Res 2000;79:474.
76. Chaves ES, Rocha EF, Marcantonio RAC, Cirelli JA, Rossa C Jr,  Marcantonio JR. Treatment of Class II furcation defects with Emdogain. An experimental study in dogs [abstr 3456]. J Dent Res 2000;79:575.