j Periodontol • January 2001 51 A Comparative Study Utilizing Open Flap Debridement With and Without Enamel Matrix Derivative in the Treatment of Periodontal Intrabony Defects: A 12-Month Re-Entry Study Stuart J. Froum,* Mea A. Weinberg/ Edwin Rosenberg,* and Dennis Tarnow* Background: Previous studies have demonstrated that enamel matrix derivative (EMD) has the ability to improve clinical parameters when used to treat intraosseous defects. The purpose of the present study was to compare at 12 months postsurgery sites treated with open flap debridement (OFD) alone to those treated with OFD and EMD. Methods: Twenty-three subjects with at least 2 intrabony defects were chosen. Fifty-three defects received EMD in conjunction with OFD. Thirtyone defects in these same 23 subjects were treated with OFD alone. Stents were fabricated to serve as fixed reference points, Re-entries were performed at least 1 year after initial surgery. Soft tissue measurements were recorded prior to initial surgery and prior to re-entry for gingiva! (GI) and plaque (PI) indices, probing depth (PD), gingival margin position, and clinical attachment level (CAL). Hard tissue measurements were recorded during the initial and re-entry surgery for level of crestal bone and depth of defect. Statistical analysis was conducted using the method of generalized estimating equations to determine changes in Gl, PI, PD, CAL, fill of the osseous defect, and crestal resorption. Percent of defect fill was also calculated. Results: In all categories, treatment with EMD (test) was superior to treatment without EMD (control). Average PI and Gl were not significantly different either initially or prior to re-entry. The average PD reduction was 2.7 mm greater with EMD than controls. The average CAL gains were 1.5 mm greater, and the average fill of osseous defect 2.4 mm greater with EMD than controls. The average percent of defect fill after adjusting for crestal bone loss was more than 3 times greater for EMD versus control-treated sites (74% defect fill with EMD versus 23% defect fill for control sites). Conclusions: This study indicates that treatment of periodontal intraosseous defects with EMD is clinically superior to treatment without EMD (open flap debridement) in every parameter evaluated. Re-entry data demonstrate that percent fill of osseous defects treated with EMD compares favorably with the treatment results utilizing bone grafts or membrane barriers, according to published literature. J Periodonlol 2001;72:25-34. KEY WORDS Periodontal regeneration; intrabony defects; clinical trials; comparison studies; follow-up studies; proteins, enamel matrix; surgical flaps. * Division of Surgical Sciences, Department of Implant Dentistry, Mew York University, College of Dentistry, New York, NY. + Division of Surgical Sciences. * Department of Implant Dentistry, Mew York University, College of Dentistry. The goals of periodontal therapy include arresting the disease process, preventing disease recurrence, and regenerating lost periodontium. The latter, which is defined as reconstruction of a functionally oriented periodontal ligament inserting in new alveolar bone and cementum, can only be determined by histologic examination of the healed tissues following surgery. Human evidence of periodontal regeneration has been presented in several literature reviews.1 6 However, since this evidence requires human block section and is therefore limited, clinicians rely on parameters such as probing depth (PD) reduction, clinical attachment level (CAL) gain, and radiographic and re-entry evidence of fill of the osseous defect (OF) to clinically evaluate a treatment modality. Several thorough literature reviews have documented success using various graft and barrier materials in achieving statistically superior clinical results compared to open flap debridement.1'7 A recent 25 Treatment of Intrabony Defects With Enamel Matrix Derivative Volume 72 • Number I meta-analysis demonstrated clinically improved PD reduction, CAL gain, and OF following treatment of intrabony defects with membranes compared to grafts, and with both materials compared to open flap debridement. 8 In 1997 enamel matrix dervative^ was introduced into the periodontal literature and has demonstrated the potential to mediate periodontal regeneration in both human and animal model systems.911 This material is composed primarily of amelogenin and related proteins that are derived from porcine tooth buds. The detection of enamel matrix proteins between the peripheral dentin and the developing cementum provides the basic concept for use of an EMD material in regenerative therapy.12 The use of EMD has been shown to result in significant clinical improvement in PD reduction, CAL gain, and bone fill in several recent clinical studies.13'15 A multicenter study involving 33 subjects with 34 paired test and control sites compared the long-term effect of EMD treatment as an adjunct to modified Wid man flap surgery to the same surgical procedure with a placebo. Clinical attachment measurements and radiographic assessments were performed at baseline and 8, 16, and 36 months postsurgery. No re-entry surgeries were done. Results showed significantly greater gains in attachment at all measurement periods in the EMD-treated sites compared to the controls. Clinical attachment gain in test and control sites at 8 months was 2.1 mm and 1.5 mm, respectively; at 16 months, 2.3 mm and 1.7 mm, respectively; and at 36 months, 2.2 mm and 1.7 mm. respectively. There was a statistically significant radiographic bone gain of 2.6 mm at 36 months at test sites, compared to no evidence of bone gain at control sites.13 A recent clinical study in 108 patients included 145 angular interproximal bone defects with a >3 mm intrabony component that were treated with EMD and reexamined 12 months following surgery. Results reported mean CAL gain of 4.6 mm and a PD reduction of 5.2 mm. Eighty-seven percent of treated sites exhibited a CAL gain >2 mm. Moreover, radiographic assessment revealed an average decrease in depth of bone defect of 2.9 mm. However, no control treatment was performed in this study.14 A recent randomized controlled clinical trial compared 4 treatment groups of 10 subjects each. Three different barriers (2 bioabsorbable and 1 nonabsorbable) were evaluated in 3 groups of 10 patients with intrabony defects. The fourth group received EMD. Controls were treated with the EMD vehicle (propylene glycol alginate) only following open flap debridement. Re-examination 12 months after surgery showed a significantly greater PD reduction (4.4 mm versus 3.5 mm) and CAL gain (2.9 mm versus 1.8 mm) in the barrier and EMD-treated sites compared to the control. Moreover, the EMD treatment appeared to be equally effective in terms of PD reduction and CAL gain as the barrier treatment.15 Another controlled clinical trial compared the treatment of intrabony defects with EMD to that with guided tissue regeneration (GTR) utilizing a bioabsorbable membrane. Sixteen patients with contralateral intrabony defects in the same jaw were treated and sites evaluated prior to surgery and 8 months later. Sites treated with EMD demonstrated a PD reduction of 3.8 mm and a CAL gain of 3.1 mm. Sites treated with GTR demonstrated a PD reduction of 4.0 and a CAL gain of 3.0 mm. The authors noted membrane exposure in 7 out of 16 cases within the first 3 postsurgical weeks. The authors concluded that "no statistically significant differences in any of the investigated parameters were observed between the EMD and GTR groups."15 In a human histological study, 14 teeth in 14 patients scheduled for extraction and associated with advanced intrabony defects were randomly treated with EMD or bioabsorbable membrane. Six-month histological analysis revealed 2.6 mm of new attachment (new cementum with inserting collagen fibers) and a mean 0.9 mm gain of new bone in the EMD-treated sites. In the GTR group, mean new attachment gain was 2.4 mm and mean gain of new bone was 2.1 mm. The authors noted that the new cementum was predominantly cellular in both treatment groups. The authors concluded that treatment with EMD or with a bioabsorbable membrane enhanced the formation of a new connective tissue attachment in humans.17 To date there has been no report in the literature of a controlled clinical trial with re-entry data utilizing EMD. The purpose of the current investigation was to clinically compare the 1 -year results of the treatment of intraosseous defects with either open flap debridement (OFD) alone or in conjunction with the use of EMD. MATERIALS AND METHODS Twenty-three patients were selected from those requiring periodontal treatment who presented to the New York University College of Dentistry. Each patient had at least 2 angular osseous defects as detected on radiographs and had completed full-mouth scaling and root planing, occlusal adjustment where indicated, and home care instructions at least 2 months prior to screening for participation in the present study. All patients had undergone initial head and neck oral cancer examination. No patients required antibiotic premedication or had any systemic disease which contraindicated periodontal surgery. Each patient was told the nature of the study and signed an informed consent prior to any treatment. Both the study and the ij Emdogain, Biora, Chicago, IL. Periodoncol • January 2001 Froum. Weinberg. Rosenberg, Tarnow % Defect Fill = consent were approved by the University Committee on Activities Involving Human Subjects. Prior to surgery, defects were assigned by a coin flip to receive either enamet matrix derivative (EMD) following open flap debridement or debridement alone. All defects in the treated quadrant received the same treatment during surgery. If three quadrants contained study defects, a blind draw determined the control defect and all defects in the remaining 2 quadrants were treated by OFD with EMD. If 4 quadrants contained treatment defects, a blind draw again determined the 2 EMD-treated quadrants and the 2 control-treated quadrants. Measurements Prior to surgery, a customized acrylic stent was fabricated for each patient and stored on the study cast to minimize distortion. The stent was grooved in an occlusal apical direction with a tapered bur so that the silver point was returned to the same position for each successive measurement. Measurements for gingival recession PD. and CAL were recorded from a fixed reference point (stent) to the nearest 0.01 mm using a digital Boley gauge, silver point, and locking pliers.18 All measurements were recorded by a single investigator. The investigator performing the measurements did not perform the surgical treatment. Moreover, since the measurements were performed prior to the coin flip that determined treatment, the measurement evaluator was blinded to the treatment performed. This was also true of the measurements recorded at re-entry surgery. Local site scores were recorded for: 1) plaque index (PI);59 2) gingival index (Gl);20 3) gingival recession (Rec); 4) probing depth (PD); 5) clinical attachment level (CAL); and 6) mobility. 21 Prior to surgery (baseline), measurements were made from the stent to the gingival margin (GM), stent to the base of the pocket (BP), and stent to the cemento-enamel junction (CEJ). During surgery, measurements were made from the stent to the alveolar crest (AC), stent to the base of the osseous defect (BD), and stent to CEJ. The latter was used to check for seating of the stent. Approximately 12 months following initial surgery, prior to, and during re-entry surgery, the above measurements were repeated (Table 1). Recordings were also made for plaque index (PI) and gingival index (Gl) prior to initial surgery and reentry. Patients who smoked more than one-half pack per day were classified as "smokers," and this was documented. Fifty-three defects in 23 patients received EMD in conjunction with OFD surgery. Six of these defects were treated in 3 smokers. Thirty-one defects in the Table I. Measurement Parameter Calculations Parameter Initial probing depth (IPD) Initial osseous depth Gingival recession 2-month probing depth Probing depth reduction Gam in clinical attachment leveS Alveolar crest resorption 1 2-month osseous depth Defect fill Measurement Stent to BP minus stent to GM Stent to BD minus stent to AC Re-entry stent to GM minus baseline stent to GM Re-entry stent to BP minus ne-entry stent to GM IPD mmus 12-month PD Baseline stent to BP minus re-entry stent to BP Re-entry stent to AC minus baseline stent to AC Re-entry stent to BD minus ne-entry stent to AC Baseline stent to BD minus re-entry stent to BD Defect fill 100 initial osseous depth _ , _ . . Initial osseous depth - 12 month osseous depth x !00 i Defect Resolution = v- t- Initiat osseous depth same 23 patients were treated with OFD alone. Three of these defects were treated in the 3 smokers. Surgical Procedures and Postsurglcal Care Following measurement recording and administration of local anesthesia, all sites were treated via reflection of a full thickness mucoperiostea! flap attempting to retain all soft tissue. The exposed roots and osseous defects were debrided with hand and ultrasonic instruments. The flap debridement-only (OFD) sites then had the flaps repositioned and sutured using 4-0 silk sutures. The EMD sites were dried with nonwoven gauze. Citric acid (pH = 1) was then applied with cotton pledgets to the exposed root for 15 seconds. Care was taken to avoid the spread of the liquid to adjacent sites The arej was then irrigated with sterile saline or water for 60 seconds. In all cases suturing, using the same 4-0 silk sutures, was begun prior to EMD application to allow more rapid closure of the site once the EMD was applied. The needle and suture were passed through the buccal flap and the flap reflected with an elevator. The root and defect were again dried and EMD applied to both the root and the defect. The EMD was prepared by mixing the 30 mg powder with the propylene glycol alginate vehicle using a 3 cc syringe^ with an 18-gauge, 2-inch needle.* Preparation of this mix was made at least 15 minutes prior to application, Using the supplied 19- I! Ethicon Inc., Sommerville, NJ. 1| Monoject Luer Lock syringe, Sherwood Medical. St. Louis, MO. # Luer Lock hypodermic needle. Air-Tite Products Co.. Inc.. Virginia Beach. VA. 27 Treatment of Intrabony Defects With Enamel Matrix Derivative Volume 72 • Number 1 gauge, 1.5-inch needle^ on the same syringe, EMD was slowly applied to the deepest part of the exposed root moving coronally until the defect was filled. The flaps were repositioned and sutured with interrupted or horizontal mattress sutures. In cases where flap coverage was judged to be incomplete, vertical releasing incisions or partial thickness periosteal releasing incisions were made at the inner portion of the base of the flap prior to root conditioning. Following suturing, additional EMD was placed at the tooth flap junction. Pressure was maintained on flaps in both experimental and control groups with damp gauze for 5 to 10 minutes. A resin dressing** was applied, covering the tissue and apical one-half of the teeth. Each patient was prescribed tetracycline HCL 1 g per day for 2 weeks following surgery. Patients were then instructed in homecare. Homecare consisted of rinsing twice a day with 0.12% chlorhexidine gluconate+t for 8 weeks. Sutures and dressings were removed 10 to 14 days postsurgery. The patients returned every 2 weeks for 6 weeks, then once a month for the next 10 months. The surgical areas were debrided professionally with supragingival scaling, and home care instructions were reinforced. Patients were instructed not to brush or floss the treated areas for 6 weeks following suture removal and then return to their normal regimen of home care. The surgical sites were not probed until 6 months postsurgery. Approximately 12 months after initial surgery, re-entry surgery was performed (Fig. 1A-D). Prior to surgery, all soft tissue measurements were repeated and recorded. At the time of re-entry, local anesthesia was administered and full thickness flaps were reflected. Defects were debrided and hard tissue measurements performed (Fig. 2A-D). Residual defects were treated with surgical elimination and the flaps sutured. Prior to initial and re-entry surgeries, radiographs ** Coe pak. GC America Inc., Alsip, IL. t + Peridex. Zila Pharmaceuticals Inc.. Phoenix. AZ. Figure I. A. Control open debndement site, mesia/ tooth #6. Presurgical probing depth measured 9.93 mm. B. Exposure of the 2-wali osseous defect measuring 5,25 mm in depth, C. At re-entry (12 months), probing depth 5.8 mm wtth CAL gain of I.58 mm. D. Remodeling of the osseous defect with on osseous fill of 1.36 mrr. Thy, represents a defect fill of 26%. 28 J Periodoncol Froum, Weinberg, Rosenberg, Tarnow and photographs were taken. Photographs of initial and re-entry hard tissue were taken of each surgical site during the surgical procedure (Fig. 3A through D). Statistical Methods The design of the study was a split-mouth design so that each subject served as his own control. In other words, there were periodontal defects within each patient that were treated without EMD and similar periodontal defects within the same patients that were treated with EMD. Data were collected initially and 12 months after surgery for mobility, probing depth, osseous depth, and plaque and gingival indices. In addition, data were collected on the change in PD over 12 months, the change in gingival recession over 12 months, the change in CAL over 12 months, the amount of crestal resorption over 12 months, the amount of defect fill over 12 months, and the percent of defect resolution over 12 months (the net gain in percent of defect fill after accounting for crestal resorption). Statistical analysis was conducted using the method of generalized estimating equations (GEE) assuming an exchangeable working correlation matrix. This method of statistical analysis was used in order to appropriately control for baseline differences among sites while at the same time taking into account the lack of independence of sites within each subject. In addition, some subjects had multiple sites treated with EMD and/or multiple sites treated without EMD. All subjects received treatment for at least one periodontal defect with EMD and for at least one defect without EMD. Descriptive Statistics The average age of the subjects in the study was 45.5 years (SD = 15.9), ranging from 19 to 71 years. Summary statistics for initial clinical parameters are given in Table 2. Using patient averages for control sites (sites without EMD) and test sites (sites with EMD), differences in initial PD, osseous depth, PI, Figure 2. A. EMD-treoted site on the mesial of tooth #30. Presurgical probing depth measured 9.96 mm. B. Exposure of the osseous defea which measured 6.35 mm. C. At re-entry (12 months), probing depth of 3.90 mm with a CAL gain of 5.46 mm. D. Fi of the osseous defea measured 4.24 mm. This represents a defect fill of 67%. 29 Treatment of Intrabony Defects With Enamel Matrix Derivative Volume 72 • Number 0L Figure 3. A. Osseous defect exposed on the distal of tooth #30 treated with EMD.The defect measured 640 mm, B. At 12-month re-entry, the defect fill measured 5.40 mm, representing a percent defect fill of 85%. C. Presurgical radiograph of the defect prior to EMD treatment. D. A ! 2-month postsurgical radiograph prior to re-entry. and Gl between control average and test average for each subject were tested using a paired t test. Initial PD, PI, and Gl were not significantly different between average test and average control for each subject (P = 0.125, P= 0.244, and P= 0.931 for initial probing depth, PI, and Gl, respectively). However, the average test osseous depth was significantly deeper than the average control osseous depth (P<0.001). Statistical Analysis Changes in PD, REC, CAL, and osseous depth over 12 months were compared using GE:E. Age and smoking status were both considered for inclusion in the model, but neither of these variables added significantly to the GEE model or changed the estimation of effect of treatment with EMD substantially, so they were eliminated from the GEE models. For comparing the change in PD over 12 months between sites treated with EMD and sites treated without EMD, initial probing depths were included as a covariate in the model. For comparing the change in recession, change in CAL, and change in osseous depth, the initial osseous depth was included as a covariate in the model. The subject was the cluster variable for conducting the GEE analyses, while an exchangeable working correlation was assumed for all the GEE analyses. In addition, the amount of defect fill, the percent defect fill, the amount of crestal bone resorption, and the percent defect resolution (unadjusted for crestal bone resorption) were also compared using GEE with initial osseous depth as a covariate. Adjusted means with 95% confidence intervals for these changes by treatment group are given in Table 3. In addition, minimum and maximum changes are also presented, in all categories of changes, treatment with 30 Periodontol • January 2001 Froum, Weinberg, Rosenberg, Tarnow Table 2. Summary Statistics for Initial Clinical Parameters by Treatment Clinical Parameter isjurrber Sites Mean SD Median Minimum Maximum Probing depth Control Test Osseous depth Control Test Plaque index Control Test Gingival index Control Test Table 3. Adjusted Mean Changes in Clinical Parameters Over 1! Months by Treatment* 31 53 31 53 31 53 31 53 7,32 7.99 4,29 5.63 0.16 0.38 0.29 0.40 1,48 1.46 0.93 1.24 0.37 0.60 0,53 0.57 7.42 7.88 4.11 5.60 0 0 0 0 4.10 5.48 2.91 2.01 0 0 0 0 1 1.44 1 1.90 7.66 8.40 1 2 2 Clinical Parameter Reduction in probing depth Contra Test Change in recession Control Test Increase in CAL Control Test Decrease in osseous depth Control Test Amount of defect fill Control Test % defect resolution Control Test Crestat bone resorption Control Test % defect fill Control Test Adjusted Mean 2.24 4.94 1,29 0.61 2.75 4.26 2.65 4.28 1.47 3.83 48.1% 83.2% 1.29 0.46 22.7% 74.0% 95% CI (1.87,2.62) (4.76.5.13) (0.99, 1.60) (0.46,0.76} (2.36,3.13) (4.03, 4.49) (2.44,2.86) (4.03,4.54) (1.17. 1.76) (3.58.4.08) (43.3. 52.8) (78.5, 87.9) (1.16, 1.43) (0.36,0.55) (17.8,27.6) (69.2,78.7) Minimum 0.00 2.43 0.05 0.00 0.28 2.43 0.00 2.01 0.00 1.05 0.0% 43.7% 0.00 0.00 0.0% 26.1% Maximum 3.85 8.50 2.65 2.00 5.38 7.63 3.54 7.01 2.52 6.20 74.6% 100.0% 2.34 1.98 51.3% 100.0% pi <0.000l 0.0001 <0.000l <0.000l <0.000l <0.000l <0.000 <0.000 Means are adjusted for baseline measures of probing depth or osseous depth. Pvalues are based en GEE models with baseline covariate included, EMD was superior to treatment without EMD. Over 12 months, treatment with EMD showed significantly greater decrease in PD, a significantly greater increase in CAL, and a significantly greater percent of fill of initial osseous defect, both with and without adjustment for crestal bone resorption. In addition, treatment with EMD demonstrated significantly less crestal bone resorption over time than treatment without EMD. Overall, the average reduction in PD when treated with EMD was 2.7 mm greater than treatment without EMD; the average gain in CAL for sites treated with EMD was 1.5 mm greater than sites treated without EMD; and the average OF of sites treated with EMD was 2.4 mm greater than sites treated without EMD. The average percent of defect fill after adjusting for crestal bone loss was over three times greater in sites treated with EMD compared to sites treated without EMD (74% with EMD and 23% without EMD). Average plaque indices and gingival indices were not significantly different upon re-entry (with EMD: mean PI = 0.51 and mean Gl = 0.71; without EMD: mean PI = 0.68 and mean GI = 0.71; P=03] and P = 1.00 for difference in PI and GI, respectively). DISCUSSION Overall, this study indicates that treatment of periodontal defects with EMD is clinically superior to treatment without EMD for each parameter measured. Since the average initial and 12-month plaque and gingival indices were not significantly different between the test (EMD) and control (no EMD), it appears that similar levels of plaque control were maintained at all treated sites. This fact is important because clinical studies utilizing open flap debridement18- 22-23 or membrane barriers24 27 have shown that the amount of PD reduction, CAL gain, and OF can be correlated to plaque control at the treated site. Although the average initial PD in the present study was not signifiTreatment of Intrabony Defects With Enamel Matrix Derivative Volume 72 • Number I cantly different between test and control-treated sites, the average initial osseous depth (IOD) was significantly deeper at the test sites. This factor has been documented in OFD procedures28 as well as in GTR treatment18242627 of intrabony defects to correlate with the postsurgical CAL gain. However, other investigators reported that defect depth was not shown to affect OFD in the treatment of intrabony defects using membrane barriers.29 Recent studies utilizing EMD to treat intraosseous defects reported conflicting results regarding the relationship of CAL gain following surgery and IOD of the defects treated. One study reported that "clinical improvements were better at sites with deep, than at sites with shallow, intrabony defects;"15 while in a second study using a regression model to analyze CAL change, it was found that the depth of the intrabony defect did not significantly influence CAL change following EMD treatment.14 While the significantly greater OF of the test versus control-treated sites in the present study may be related to the greater initial depth of defect in the test group (5.63 versus 4.29), other defect characteristics may also have affected the fill. The number of osseous walls30 and width of the defect (angle of the defect)31 are 2 other factors that have been shown to correlate with OF. These parameters were not measured in this study. However, since defect treatment was selected by coin flip and no measurements were recorded for the width, height, or volume of the treated defects, some, none, or all of these factors may have equally affected the results of both the test and control groups. Moreover, the inclusion of initial osseous depth as a covariate in our statistical model adjusted for difference in response related to differences in the size of the initial defects and thus eliminated any bias. Therefore, the clinical significance of the difference in fill between test and control-treated sites in light of the above factors is evident, with all the results demonstrating clear superiority of the test group over the control group. To date, our results are consistent with other controlled studies comparing EMD to open debridement which have demonstrated greater OF of the EMD-treated sites.1315 Nevertheless, OF of EMD treated sites in the present study, as determined by re-entry measurement, was 3.83 mm which exceeded the OF reported by both Heden et al.14 (2.8 mm) and Heijl et a!.13 (2.6 mm) (where fill was determined by radiographs). Furthermore, the percent of defect fill in the present study was 74% with EMD and 22.7% with OFD. This calculation is a ratio of the OF and the initial defect depth which serves to standardize the bone fill observed. The CAL gain and PD reduction of 4.26 mm and 4.94 mm, respectively, in EMD-treated sites in this study also compare favorably to that reported by Heden et al.14 (4.6 mm and 5.2 mm) and Pontoriero et at.15 (2.9 mm and 4.4 mm) at intraosseous sites treated with EMD. It is interesting to note that in the present study, no treated site either in the test or control group showed a clinical loss of attachment. Moreover, with the exception of one site in the control group, which showed no gain or loss in fill of the osseous defect, all other sites in the 2 treated groups showed a positive OF. The reasons for the positive results seen in the present patient population may include the level of plaque control and the frequency of professional maintenance which, in a previous study utilizing OFD, was shown to directly correlate with CAL gain and OF.18 The clinically superior results seen in this study in the sites treated with EMD compared to control sites may be indicative of the biological activity of EMD proteins, which are involved in the differentiation of cementoblasts. Increasing evidence shows that Hertwig's epithelial root sheath (HERS) plays an important role in both cellular and acellular cementum formation.32 Hammarstrom9 demonstrated that cells of HERS synthesize enamel-like substances termed enamel matrix proteins (EMPs). The deposition of EMPs including the amelogenins, the enamelins, and the amelins on the root surface33 may precede the formation of acellular cementum,9-32 providing an essential surface for the expression of cementoblasts. Once cementum has been laid down on top of the enamel matrix-covered dentin surface, a series of events occur leading to the generation of the periodontal attachment apparatus. Thus, in addition to an exposed dentin surface, HERS or its products may participate in the formation of acellular cementum.34 Cells close to the root surface are not only involved in cementum formation but also in the formation of periodontal ligament and alveolar bone.35 Enamel matrix derivative is composed primarily of amelogenin and other related proteins that are derived from porcine tooth buds.9 Thus, the adjunctive use of enamel proteins in conjunction with regenerative periodontal surgery may provide a natural extracellular matrix for recolonization of previously diseased root surfaces with cells expressing a cementoblast phenotype. l0 Since the body is normally exposed to the proteins of the enamel matrix during tooth development, these proteins are probably recognized as "self" by the immune system so that there have been no reports of rejection.3637 A recent study38 documented that EMPs have a stimulatory effect on the attachment and spread of human periodontal ligament fibroblasts (HPLF) on the root surface during the early stages of wound healing when compared with human gingival fibroblasts (HGF). EMPs also stimulated the expression of alkaline phosphatase which may increase the cementogenic capacity of HPLF. Additionally, EMPs were found to stimulate the release of transforming growth factor-pi (TGF-pl) by HPLF and HGF. It was speculated that J2 J Periodontoi • January 2001 Froum, Weinberg, Rosenberg, Tarnow this difference in attachment rate between the 2 fibroblast populations is related to differences in the expression of certain membrane-bound adhesion receptors. The difference in fibroblast migration rates as well as the ability of EMP to stimulate TGF-fil production may explain its contribution to selective cell repopulation of treated defects leading to the clinical results documented in the present study. Materials or techniques are often compared by analyzing the results of statistics from studies that utilize only one of the materials. An example would be a recent series of case reports utilizing EMD that showed a CAL gain of 4.6 mm and PD reduction of 5.2 mm.l 4 Based on these findings, comparisons were made to the results in a published meta-analysis of 16 clinical studies and 545 defects which were treated with GTR. bone grafts, or OFD procedures.8 The authors of the former study concluded that "... the results following Emdogain therapy, is similar to the corresponding outcome variables following GTR." Conclusions drawn from these comparisons are often fraught with error because of the difference in patient population, measurement techniques, surgical approach, operator skill with a given material, type of defect treated, and levels of plaque control. We have seen the effect of only one of these variables, operator differences, in a recent paper on GTR in the treatment of intrabony defects in a multicenter study.39 Treatment findings among the various centers showed a 1.73 mm difference in CAL gain, with 3 of 1 1 centers showing a greater gain in CAL with the OFD control compared to the GTR-treated sites. Therefore, the results presented in the current study only contrast the results of EMDtreated sites compared to the OFD control. However, as mentioned earlier, 2 previously published studies,l5'16 with 12- and 8-month postsurgical evaluations, respectively, demonstrated no statistically significant differences in improvements of PD reduction and CAL gain with EMD compared to membrane-treated sites. Before any equivalency conclusions are made, larger sample sizes and more studies are required.40 Moreover, since all defects in the above-mentioned as welt as the current investigations were intraosseous-type defects, extrapolation of these results to other defect types {i.e., furcations, horizontal bone loss, buccal plate defects) would be inappropriate. Lastly, a recently published editorial discussed the distinction between statistically significant and clinically significant results.4' Based on the PD reduction (4.94 mm), CAL gain (4.26 mm), OF (4.28 mm), percent of defect fill (74%), and percent of defect resolution (83.2%) which were obtained in sites in this study 1 year following treatment with EMD, few practitioners would argue the clinical significance of these results. This certainly adds an additional clinical modality for treatment of intraosseous defects. ACKNOWLEDGMENTS The authors would like to thank Dr. Martha Nunn, Department of Public Health Sciences, Baylor College of Dentistry, Dallas, Texas, for her invaluable help with the preparation and analysis of the statistics for the study. REFERENCES 1. Hancock EB. Regenerative procedures. In: Proceedings of the World Workshop in Clinical Periodontics. Chicago: The American Academy of Periodontology; 1989:VI1- VI26. 2. Froum SJ, Gomez C. Periodonta! regeneration. CurrOpin Periodonlol 1993; 1:1 1 1-128. 3. Becker W, Becker BE. Periodontal regeneration updated. JAm Dent Assoc 1993;214:37-43. 4. Garrett S, Bogle G. Periodontal regeneration. Curr Opin Periodontoi 1994;2:168-177. 5. Brunsvold MA, Mellonig JT. Bone grafts and periodontal regeneration. Periodontoi 2000 1993; 1:80-91. 6. Garrett S. Periodontal regeneration around natural teeth. Ann Periodontoi 1996; 1:621 -666. 7. Egelberg J. Regenerative treatment of intraosseous defect in periodontics. In: The Scientific Way, 3rd ed. Malmo, Sweden: Odonto Science; 1999:267006. 8. Laurell L, Gottlow J. Zybutz M, Persson R. Treatment of intrabony defects by different surgical procedures. A literature review. J Periodontoi 1998;69:303-313. 9. Hammarstrom L. Enamel matrix and cementum development, repair and regeneration. J Clin Periodonlol 1997;24:658-668. 10. Heijl L. Periodontal regeneration with enamel matrix derivative in one human experimental defect. A case report. J Clin Periodontal 1997;24:697-704. 1 I. 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. 12. Hammarstrom L, Heijl K, Gestrelius S. Periodontal regeneration in a buccal dehiscence model in monkeys after application of enamel matrix proteins. J Clin Periodontoi 1997;24:669-677. 13. Heijl L, Heden G, Svardstrom G, Ostren A. Enamel matrix derivative (EMD) in the treatment of intrabony periodontal defects. J Clin Periodontoi I997;24:7O5-714. 14. Heden G, Wennstrom J, Lindhe J. Periodontal tissue alterations following EMD treatment of periodontal sites with angular bone defects. A series of case reports. J Clin Periodontoi 1999;26:855-860. 15. Pontoriero R, Wennstrom 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 Periodontoi 1999;26:833-840. 16. Sculean A, Donos M, Blaes A, Lauermann M, Reich E, Brecx M. Comparison of enamel matrix proteins and bioabsorbable membranes in the treatment of intrabony periodontal defects. A split mouth study. J Periodontoi 1999:70:255-262. 17. Sculean A, Donos M, Windisch P, et al. Healing of human intrabony defects following treatment with enamel matrix proteins or guided tissue regeneration. J Periodont Res 1999:34:310-322. 1 8. Froum SJ, Coran M, Thaller, Kushner L, Scopp IW, Stahl SS. Periodontoi healing following open debridement flap procedures. J Periodontoi 1982:53:8-14. Treatment of Intrabony Defects With Enamel Matrix Derivative Volume 72 • Number I 19. Silness J. Loe H. Periodontal disease in pregnancy. II. Correlation between oral hygiene and periodontal condition. Ada Odontol Scand 1964;22:121-135. 20. Loe H, The gingival index, the plaque index and the retention index systems. (Part II). J Periodonlol 1967; 38:610-616. 21. Miller SC. Textbook of Periodontia, 3rd ed. Philadelphia: The Blakiston Company; 1950:125. 22. Myman S, Lindhe J, Rosling B. Periodontal surgery in plaque infected dentitions. J Clin Periodonlot 1977;4:240- 249. 23. Rosling B, Nyman S, Lindhe J The effect of systematic plaque control on bone regeneration in infrabony pockets. JCtin Periodontol 1976:3:38-53. 24. Tonetti M. Pini Prato G, Cortellini P. Periodontal regeneration of human infrabony de-ects. IV. Determinants of the healing response. J Periodontol 1993:64:934-940. 25. Cortellini P, Pini Prato G, Tonetti M. Periodontal regeneration of human infrabony defects. V. Effect of oral hygiene on long-term stability. J Clin Periodontol 1994; 21:606 610. 26. Machtei EE, Cho Ml, Dunford R, et al. Clinical, microbiological, and histological factors which influence the success of regenerative periodontal therapy. J Periodonto! I994;65:154-16 I. 27. Tonetti M, Pini Prato Q, 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. 28. Renvert S, Garrett S. Nilveus R, Chamberlain ADH, Egelberg J. Healing after treatment of periodontal intraosseous defects. VI. Factors influencing the healing response. J Clin Periodontol 1985; 12:707-715. 29. Trombelli L, Kim CK, Zimmerman GJ. Wikesjo UME. Retrospective analysis of factors related to clinical outcome of guided tissue regeneration procedures in intrabony defects. J Clin Periodontol 1997:24:366-371. 30. Ellegaard B. Loe H. Mew attachment of periodontal tissues after treatment of intrabony lesions. J Periodontol 1971:42:648-652. 3 I. Steffensen B, Weber HP. Relationship between the radiologic periodontal defect angle and healing after treatment. J Periodontol 1989;60:248-254. 32. Hammarstrom L. Alatli I, Fong CD. Origins of cementum. Oral Dis ! 996:2:63-69. 33. Slavkin HC. Towards a cellular and molecular understanding of periodontics: Cementogenesis revisited. J Periodontol 1976;47:249-255. 34. Thomas HF. Kollar EJ. Tissue interactions in normal murine root development. In: Davidovitch A, ed. The Biological Mechanisms of Tooth Eruption and Root Development. Birmingham, AL: EBSCEO Media; 1988:145- 151. 35. Ten Cate AR, Mills C, Solomon G. The development of the periodontium. A transplantation and autoradiographic study. Anal Rec 1971; 170:365-379. 36. Zetterstrom O, Andersson C, Eriksson L, et al. Clinical safety of enamel matrix derivative (Emdogain) in the treatment of periodontal defects. J Clin Periodonlol 1997; 24:697-704. 37. Wilson TG Jr. Safety testing of Emdogain. In: Wilson TG Jr, ed. Periodontal Regeneration Enhanced-Clinical Applications of Enamel Matrix Proteins. Chicago: Quintessence Publishing Co., Inc.; 1999:23-25. 38. 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 (il release of periodontal ligament and gingival fibroblasts. J Periodonlol 2000;71:31 -43. 39. Tonetti MS, Cortellini P, Suvan JE. et al. Generalizability of the added benefits of guided tissue regeneration in the treatment of deep intrabony defects. Evaluation in a multi-center randomized controlled clinical trial. J Periodontot 1998;69:I 183-1192. 40. Gunsolley J. Equivalence, superiority and negative clinical trials (guest editorial). J Periodontol 1998:69:608. 4 1. Rethman MP, Nunn ME. Clinical versus statistical significance (guest editorial). J Periodontot 1999;70:700- 702. Send reprint requests to: Dr. Stuart Froum, 17 West 54th Street. Suite 1 C/D, Mew York, MY 10019. Accepted for publication June 5, 2000. 34