J Clin Periodontol 2000; 27: 333–340 Copyright C Munksgaard 2000Printed in Denmark . All rights reserved

ISSN 0303-6979

Sofia Karapataki 1,2,Anders Hugoson 1,2, Hanne Falk 1,Healing following GTR treatmentLars Laurell 1 and Carl F. Kugelberg 3

1Department of Periodontology, The Institutefor Postgraduate Dental Education,of intrabony defects distal to Jonkoping, 2Department of Periodontology,Faculty of Odontology, Goteborg University,Goteborg, 3Department of Oral Surgery,County Hospital, Kalmar, Swedenmandibular 2nd molars using

resorbable and non-resorbablebarriersKarapataki S, Hugoson A, Falk H, Laurell L, Kugelberg CF: Healing followingGTR treatment of intrabony defects distal to mandibular 2nd molars using resorbableand non-resorbable barriers. J Clin Periodontol 2000; 27: 333–340. C Munks-gaard, 2000.

AbstractAims: The objectives of the present, randomised clinical trial were (i) to evaluatethe healing of periodontal intrabony defects at the distal aspect of mandibular2nd molars using a resorbable polylactic acid (PLA) barrier and a non-resorbablepolytetrafluoroethylene (e-PTFE) barrier and (ii) to compare the therapeutic effectof the bioresorbable versus the non-resorbable barrier.Method: 19 patients with intrabony defects distal to mandibular 2nd molars >4mm (on radiographs) were included in the study. The defects all remained 5 yearsafter surgical removal of impacted 3rd molars. Following flap elevation and defectdebridement, the defects were randomly covered with, either a resorbable PLAor a non-resorbable e-PTFE barrier. Flaps were repositioned and sutured to com-pletely cover the barriers. Treatment was evaluated clinically after 1 year by meas-urements of probing depth (PD), probing attachment level (PAL), and probingbone level (PBL) and radiographically by measurements of bone levels on com-puter digitised images of radiographs taken immediately before and 1 year post-surgery.Results: Both treatments resulted in significant PD reduction, PAL gain, and bonefill. The total PD reduction was 5.3∫l.9 mm for the PLA treated sites and 3.7∫l.7mm for the e-PTFE treated sites (p∞0.05). The corresponding values for PALgain were 4.7∫0.7 mm and 3.6∫1.7 mm (p∞0.05) and for PBL gain 5.1∫1.2and 3.3∫2.0 mm (p∞0.05). Radiographic bone fill averaged 3.4∫l.2 for the PLAand 2.0∫1.6 mm for the e-PTFE barriers (p∞0.05). Radiographic bone level meas-urements were significantly smaller than the corresponding clinical measure-ments, indicating that radiographs tend to underestimate bone fill.

Key words: guided tissue regeneration;Conclusions: GTR treatment of deep intrabony defects distal to mandibular sec-healing; barriers; intrabony defects;ond molars using resorbable PLA barriers resulted in significant PD reduction, resorbable; non-resorbable

PAL gain and bone fill at least equivalent to the results obtained using non-resorbable e-PTFE barriers. Accepted for publication 28 June 1999

A number of studies on the prevalenceof impacted third molars (M3s) havebeen presented in the literature (for re-view, see Hugoson & Kugelberg (1987)).

According to the conclusion of the Con-sensus Development Conference on Re-moval of Third Molars (National Insti-tute of Health, Bethesda, Md., USA)

held in 1979, impacted M3s should beremoved when there is evidence of path-ology. Some kind of pathology can beexpected in 12% of an impacted M3

334 Karapataki et al.

population, which corresponds toaround 2% of the general population(Stanley et al. 1988).

The least traumatic way to remove acompletely or partially impacted M3 isto reflect a flap, remove the bone sur-rounding the tooth, and divide thetooth into sections which are removedseparately. Clinical investigations, how-ever, have shown residual periodontalintrabony defects at the distal aspect ofthe second molar (M2) after surgicalremoval of impacted lower M3s (Gron-dahl & Lekholm 1973, Chin Quee etal. 1985, Marmary et al. 1985, Kugel-berg et al. 1985, Kugelberg 1990). In aretrospective study comprising 215 pa-tients, Kugelberg et al. (1985) foundthat 2 years after surgery, the preva-lence of intrabony defects exceeding 4mm was 43%. Kugelberg (1990) alsofound that the strongest predictor ofsuch complications was the age of thepatient at the time of surgery. Theprevalence of intrabony defects 4 yearsafter surgery was approximately 4% inpatients 25 years or younger and 44%in patients more than 25 years old atthe time of surgery.

Pecora et al. (1993) evaluated healingfollowing surgical removal of impactedM3s. At the time of M3 removal, test de-fects were covered with a non-resorbablee-PTFE barrier before the flaps were re-positioned and sutured, while the con-trol sites were treated in a conventionalmanner. Treatment was evaluated after 1year. Sites treated by traditionalmethods showed significant residual de-fects distal to the M2s whereas sitestreated with a barrier did not. Evidently,according to the results from this study,periodontal lesions distal to M2 can besuccessfully treated by a GTR procedureif diagnosed at the time of M3 removal,thereby preventing the later appearanceof residual, usually progressive, peri-odontal intraosseous defects distal to theM2. In a study by Karapataki et al.(1998), however, these results could notbe verified: In an intraindividual com-parison no differences in healing be-tween sites treated with a PLA barrierand sites traditionally treated werefound since most sites healed very wellirrespective of treatment. They thereforeconcluded that whether or not peri-odontal defects will persist after surgicalremoval of impacted third molars is dif-ficult to predict and suggested that suchareas should be evaluated and if necess-ary treated at a later stage.

The aim of the present study was to

evaluate the healing of periodontal in-trabony defects at the distal aspect ofmandibular second molars using re-sorbable PLA barriers and non-re-sorbable e-PTFE barriers.

Material and Methods

The study was designed as a random-ised study, initially comprising 20 pa-tients with a mean age of 43∫7 years.The patients included in the study hadbeen recruited from the Department ofOral Surgery at the Institute for Post-graduate Dental Education, Jonkoping,Sweden, on the basis of having a peri-odontal intrabony defect distal to amandibular second molar (M2) 5 yearsor more after surgical removal of theadjacent impacted third molar (M3).None of the patients suffered from peri-odontitis or exhibited general peri-odontal break down.

Inclusion criteria

To enter the study the patient was re-quired to fulfil the following criteria:O full mouth plaque scoreΩ20%;O presence of an intrabony defect distal

to M2Ω4 mm as measured on radio-graphs;

O probing attachment level distal toM2Ω6 mm;

O no restorations or carious lesions onroot surfaces associated with the in-trabony lesion.

Pre-experimental treatment

Following patient selection and in-formed consent, each patient was giveninitial periodontal therapy on an indi-vidual basis including instruction inproper oral hygiene techniques, scaling,and root planing. During a 1-monthfollow-up period, additional instructionand re-enforcement of oral hygiene wasprovided according to individual needs.

Periodontal surgery

The area selected for surgery was anaes-thetised with XylocainA adrenaline1:100,000 (Astra Sweden). Starting 20mm distal to the M2, buccal and lingualintracrevicular incisions extending to atleast the mesio-buccal/lingual line angleof the first molar were made. The in-cisions were terminated buccally by avertical releasing incision extendinginto the alveolar mucosa. Full thickness(muco-periosteal) flaps were raised on

the buccal and lingual aspects of theteeth and extended to achieve properaccess to the defect.

The exposed defect was then meticu-lously debrided and the root surfacesscaled and planed. To enhance bone re-generation, decorticating or intra-mar-row penetration through the corticalwall of the defect with a 2 mm roundburr was performed. During these pro-cedures, the area was generously irri-gated with sterile saline. After the de-fect had been prepared, a randomis-ation envelope was opened containinginformation whether the site was to becovered with a resorbable PLA barrier(GUIDORA bioresorbable matrix bar-rier), or a non-resorbable e-PTFE bar-rier (GORE-TEX Periodontal Ma-terialA). The barriers were applied ac-cording to the instructions provided bythe manufacturers and the flaps werecoronally positioned and sutured tocompletely cover the barrier.

Postsurgical care

The patients were given penicillin-V 1 g3¿ daily for 5 days to prevent post-operative infection and advised to re-frain from tooth brushing in the treatedarea for 6–7 weeks. In addition, to in-hibit the formation of plaque, the pa-tients were advised to rinse with a 0.2%solution of chlorhexidine digluconate(HibitanA Dental, ICI, Sweden) twicedaily for 6–7 weeks. Analgesics (Par-acetamol 500 mg, 2 tablets), were givenimmediately after surgery and then 1tablet every 3 h during the first dayafter surgery. After that analgesics weretaken only as indicated. Flap sutureswere removed after 2–3 weeks.

Surgical removal of e-PTFE barriers

The e-PTFE barrier was removed 6weeks after the first surgery. An incisionwas made through the tissue coveringthe barrier and a flap was raised to ex-pose it. The barrier was dissected freeand carefully removed. Remnants ofepithelium on the inner side of the flapwere removed. It was then repositionedto cover the newly formed tissue andsutured. After 1 week, the sutures wereremoved and rinsing with chlorhexidinesolution was discontinued.

Maintenance care

Mechanical tooth cleaning was resumed6–8 weeks after surgery. Consistent with

GTR treatment of intrabony defects 335

conventional postsurgical care, all pa-tients were recalled for plaque scoringat 4-week intervals, reinstructed in oralhygiene as needed, and subjected to‘‘professional tooth cleaning’’ (Axels-son & Lindhe 1974). This was con-tinued for 3 months postsurgery. There-after the patients were seen at 6 and 12months. The final examination was per-formed 12 months postsurgery.

Assessments

Clinical measurements were made atthe mid-buccal, disto-buccal, distal,disto-lingual, and mid-lingual aspectsof M2. Only the deepest sites weremeasured. The following registrationswere made at the baseline and l-yearexaminations:1. oral hygiene (plaque index, PLI, Sil-

ness & Loe 1964);2. soft tissue index (STI), i.e., to evalu-

ate gingival and mucosal tissuescovering the defect (index 0–3; modi-fied from Demolon et al. 1993);0Ωhealthy, non-inflamed soft tissue

at the barrier covered area;1Ωinflamed gingival margin but

predominantly healthy soft tissuecovering the barrier;

2Ωgeneral redness of the tissuecovering the barrier but no swell-ing and/or suppuration;

3Ωgeneral redness, swelling, and/orsuppuration.

3. probing depth (PD);4. probing attachment level (PAL)

measured from the cemento-enameljunction (CEJ) or other fixed refer-ence point to the bottom of the de-fect using a calibrated periodontalprobe (f: 0.5 mm); the measurementswere made to the nearest 1 mmoriented along the long axis of thetooth.Probing bone level (PBL) measuredfrom the CEJ or other fixed referencepoint to the bottom of the defectwhile the patient was under anaes-thesia.In addition, examinations for oral

hygiene and STI were performed at 2weeks and at 1, 2, 3, and 6 months aftersurgery.

Clinical examination during surgery

After debridement of the bone defects,the following measurements were madeto the nearest l mm with the same 1-mm calibrated periodontal probe:1. distance from the CEJ (or a restora-

tion margin) to the bottom of the in-trabony defect (clinical ABL);

2. distance from the bone crest to thebottom of the intrabony defect (clin-ical IBD);

3. width of the bone defect from theroot surface to the bone crest.

Presurgical (baseline) and 1-yearmeasurements were made by either oftwo calibrated examiners indepen-dently, whereas measurements duringsurgery and at the intermediate follow-up were made by the one performingthe surgery.

Radiographic examination

With the use of commercially availablefilm holder devices designed to holdvertically-oriented X-ray film, stan-dardised radiographs of the teeth in-cluded in the study were taken using theparallel technique immediately beforeand l year after therapy. Each film hold-er device was adapted to the occlusalsurfaces of the experimental tooth re-gion with the aid of an individual oc-clusal impression index (PresidentA).For the analysis of the radiographs, lin-ear measurements were made on com-puter digitised images of the radio-graphs using the digital X-ray pro-cessing and analysis method describedby Jeffcoat et al. (1984) and Tonetti etal. (1993). The analysis was performedat the University of Washington Re-gional Clinical Dental Research CenterImage Laboratory in Seattle, Wash.,USA.The following linear distances were as-sessed:O distance from the CEJ (or a restora-

tion margin) to the bottom of the in-trabony defect (radiographic ABL);

O distance from the bone crest to thebottom of the intrabony defect(radiographic IBD);

O distance from the CEJ (or a restora-tion margin) to the bone crest.

These linear measurements were ex-pressed in pixels and converted to mm.The amount of bone fill was assessed bysubtracting the pixel value of radio-graphic ABL, at 12 months from thatat baseline and then converting the re-mainder to mm. Defect resolution wasobtained by comparing the two radio-graphic IBD values. The distances fromthe CEJ to the bone crest at the twoexaminations were compared to evalu-ate crestal resorption.

For each variable and time point,double measurements were made and

the mean value of the 2 measurementswas used for defect characterisationand for further calculations.

To evaluate the precision of themethod, the differences between theduplicate measurements of the threevariables root length and radiographicABL and IBD were assessed. The pixelvalues were then converted to mm andthe error of the method expressed as thestandard deviation of the difference be-tween the 2 measurements. The stan-dard deviations amounted to 0.2 mmfor root length and 0.3 mm for bothABL and IBD. Simple regression analy-ses found the correlation between themeasurements to be 0.99 for rootlength, 0.98 for ABL, and 0.94 for IBD.

Statistical analysis

The location of the intrabony defectwith the largest IBD value as measuredat five measurement points duringsurgery (and usually located mid-dist-ally) was selected for data analysis. In-dividual as well as mean and standarddeviation values were presented. Thepaired t-test was used to evaluate differ-ences between baseline and end-pointdata of continuous variables (PD, PAL,PBL, and radiographic ABL and IBD).Differences in outcome between treat-ments were tested for statistical signifi-cance using ANOVA. In case baselinedata between groups differed, an AN-COVA test using the baseline data ofthe particular variable as a covariatewas performed to find out whether thebaseline value interacted with the treat-ment outcome. When the requirementsfor parametric tests were not met, anon-parametric Mann-Whitney U-testwas used. To test the correlation be-tween measurements, simple regressionanalyses were performed. A computerstatistical package was used in the dataanalysis (Super ANOVA, Abacus Con-cepts Inc., Berkeley, Calif., USA). Testswere two-tailed and at the 0.05 (p<0.05)significance level.

Results

20 patients participated in the study. 1patient in the e-PTFE treated group didnot show up after the 3-month appoint-ment and was excluded from analysis.Thus, the final material consisted of 10patients who had been treated withPLA barriers and nine patients whohad been treated with e-PTFE barriers.Although not consistent for all the vari-

336 Karapataki et al.

Table 1. Plaque index and soft tissue index at each site at baseline (BL), at 2 weeks, and l, 3,6 and 12 months after surgery

Plaque index Soft tissue index

Subject Treatment BL 2 w 1 m 3 m 6 m 12 m BL 2 w 1 m 3 m 6 m 12 m

LC PLA 1 0 0 0 0 1 1 0 0 0 0 1K-L J PLA 0 0 0 0 0 0 0 0 0 0 0 0L K PLA 0 0 0 1 0 1 1 0 0 0 0 0A L PLA 0 0 0 2 0 0 0 0 0 0 0 0A-E L PLA 1 0 0 0 0 0 0 0 0 0 0 0K L PLA 0 1 1 0 0 1 0 0 0 0 0 0S L PLA 0 0 0 0 0 2 0 0 0 0 0 0A S PLA 0 0 0 0 0 0 0 0 1 0 0 0B S PLA 0 0 0 0 2 2 0 2 2 0 0 0P S PLA 0 1 1 0 0 2 0 0 0 0 0 0H A e-PTFE 0 0 0 0 1 1 0 0 0 0 0 0P B e-PTFE 1 0 1 0 1 1 0 1 2 0 0 0N A e-PTFE 0 1 0 0 0 2 0 2 3 0 0 0Å J e-PTFE 1 1 2 1 2 2 0 2 2 1 1 0L-O K e-PTFE 0 0 0 0 3 0 0 0 2 0 0 0L M e-PTFE 0 0 0 1 0 2 0 0 0 0 0 0C-G M e-PTFE 0 0 0 1 1 1 1 1 0 0 0 0G P e-PTFE 0 0 0 2 0 2 0 2 3 2 0 0A N e-PTFE 0 0 0 0 0 0 2 2 2 0 0 0

Table 2. Mean and standard deviation values for probing depth, probing attachment levels,and probing bone levels at baseline and 12 months after surgery at PLA (nΩ10) and e-PTFEtreated (nΩ9) sites

Probing depth Probing attachment level Probing bone level

PLA e-PTFE PLA e-PTFE PLA e-PTFE

baseline 9.5∫1.4* 8.2∫1.1 9.6∫1.8 8.4∫0.7 10.5∫1.8* 9.0∫0.912 months 4.2∫1.4 4.6∫1.3 4.8∫1.7 4.9∫1.8 5.4∫1.2 5.7∫1.7change 5.3∫1.9* 3.7∫1.7 4.7∫0.7* 3.6∫1.7 5.2∫1.3* 3.3∫2.0p ∞0.001 ∞0.01 ∞0.001 ∞0.01 ∞0.001 ∞0.01

*Difference between treatments statistically significant; p∞0.05.

Table 3. Individual probing depths, probing attachment levels, and probing bone levels (mm)at baseline (BL) and 12 months after surgery at PLA and e-PTFE treated sites. (DiffΩdiffer-ence)

Probing attachmentProbing depth level Probing bone level

Subject Treatment BL 12 m Diff BL 12 m Diff BL 12 m Diff

C PLA 9 4 5 8 4 4 9 4 5K-L J PLA 11 4 7 12 6 6 12 6 6L K PLA 10 2 8 10 5 5 11 5 6A L PLA 8 5 3 9 4 5 9 5 4A-E L PLA 10 5 5 10 6 4 12 5 7K L PLA 7 5 2 7 3 4 7 4 3S L PLA 9 4 5 9 4 5 10 6 4A S PLA 9 4 5 9 5 4 10 5 5B S PLA 12 4 8 12 7 5 13 6 7P S PLA 10 5 5 12 7 5 12 6 6H A e-PTFE 9 6 3 9 6 3 9 6 3P B e-PTFE 6 4 2 8 4 4 9 6 3N A e-PTFE 9 3 6 9 3 6 9 3 6Å J e-PTFE 9 7 2 8 8 0 9 9 0L-O K e-PTFE 9 3 6 9 7 2 9 7 2L M e-PTFE 8 4 4 7 3 4 7 5 2C-G M e-PTFE 9 4 5 9 4 5 10 4 6G P e-PTFE 7 5 2 9 5 4 9 6 3A N e-PTFE 8 5 3 8 4 4 10 5 5

ables characterising the defects, a ten-dency towards deeper defects was foundin the PLA treated group compared tothe e-PTFE treated group.

In general, healing following bothprocedures proceeded without compli-cations or major discomfort. Loss ofsoft tissue integrity and barrier ex-posure during the first weeks of healingoccurred in 3 PLA treated sites and intwo e-PTFE treated sites. Except forone patient in the e-PTFE group withsevere flap sloughing, exposures were inthe 2–3 mm range. Infection in the sur-gical area which induced removal of thee-PTFE barrier earlier than scheduleddid not occur.

Plaque Index and Soft Tissue Index (Table1)

Postsurgical compliance was in generalof a high level with few sites showingplaque or gingivitis. Thus, during thehealing period, plaque was found in-frequently at both PLA and e-PTFEtreated sites. Inflammation in the softtissues covering the barrier, however,was found during the first month ofhealing in 2 of 10 PLA treated sitescompared to 6 of 9 e-PTFE sites. In themajority of the sites, the inflammationdisappeared after 4 weeks or after bar-rier removal.

Probing depth, probing attachment level,and probing bone level (Table 2)

Prior to surgery (baseline), mean prob-ing depths (PD) measured 9.5∫1.4 mmat PLA-treated sites and 8.2∫1.1 mm ate-PTFE sites (p∞0.05). Both treatmentsresulted 12 months later in a significantreduction in PD. Although the residualprobing depth did not differ betweenthe two groups, the PD reductiontended to be larger (p∞0.05) among thePLA treated sites (5.3∫1.9 mm) thanamong the e-PTFE treated sites(3.7∫1.7 mm). The ANCOVA showedthat initial probing depth did not inter-act with this difference in treatmentoutcome.

The probing attachment levels (PAL)were reduced from 9.6∫1.8 mm to4.8∫1.7 mm (p∞0.001) in the PLAtreated group and from 8.4∫0.7 mm to4.9∫1.8 mm (p∞0.01) in the e-PTFEtreated group. The PAL gain was4.7∫0.7 mm and 3.6∫1.7 mm respec-tively, the difference in PAL gain beingstatistically significant (p∞0.05).

Baseline PBL value was somewhat

GTR treatment of intrabony defects 337

higher in the PLA-treatment than in thee-PTFE treatment group, 10.5∫1.8 mmversus 9.0∫0.9 mm (p∞0.05). Bothtreatments resulted in significant PBLgain. Sites treated with PLA barriers re-sulted in a decrease in the probing bonelevel (PBL) from 10.5∫1.8 mm to5.4∫1.2 mm, a gain of 5.2∫1.3 mm(p∞0.001). In the e-PTFE treated sitesPBL was reduced from 9.0∫0.9 mm to5.7∫1.7 mm, a reduction of 3.3∫2.0mm (p∞0.01). The PBL gain differedsignificantly (p∞0.05) between the treat-ments. The ANCOVA test did not showany interaction between initial PBL andtreatment outcome.

Individual PD, PAL, and PBL datafor all treated sites are presented inTable 3.

Bone measures

Clinical data of PBL at baseline (Table3) and clinical ABL as measured during

Table 4. Mean and (SD) values for clinical measurements during surgery (ABL, IBD, and the width of the bone defect) and radiographicbone measurements (mm) at PLA (nΩ10) and e-PTFE treated sites (nΩ9) at baseline and 12 months after surgery

Clinical ABL Clinical IBD Defect width Radiographic Radiographic(at surgery) (at surgery) (at surgery) ABL IBD

PLA e-PTFE PLA e-PTFE PLA e-PTFE PLA e-PTFE PLA e-PTFE

Baseline 10.5∫1.8 9.5∫1.2 7.7∫1.6 6.1∫1.9 3.0∫0.7 3.8∫1.0 8.5∫1.6 8.3∫1.7 5.3∫0.9 4.7∫1.212 months 5.1∫1.7 6.3∫1.7 1.0∫1.4 1.8∫1.4Change 3.4∫1.2* 2.0∫1.6 4.4∫1.7* 3.0∫1.7p ∞0.001 ∞0.01 ∞0.001 ∞0.001

*The difference between treatments with PLA and e-PTFE is statistically significant; p∞0.05.

Table 5. Individual clinical measurements during surgery (ABL, IBD, and the width of the bone defect) and radiographic bone measurements(mm) at PLA and e-PTFE treated sites at baseline (BL) and 12 months after surgery

Clinical examination Radiographic Radiographicduring surgery ABL IBD Residual Crestal

Subject Treatment ABL IBD Width BL 12 m Diff BL 12 m Diff IBD RA

L C PLA 9 6 3 9.6 5.2 4.4 6.8 0.7 6.1 0.7 1.7K-L J PLA 11 7 3 10.0 7.8 2.2 5.3 2.7 2.7 2.7 0.4L K PLA 10 7 3 9.0 5.5 3.4 6.5 0.0 6.5 0.0 3.0A L PLA 9 7 3 7.7 4.7 3.0 6.2 0.0 6.2 0.0 3.1A-E L PLA 12 7 4 8.9 4.5 4.4 4.8 0.0 4.8 0.0 0.5K L PLA 9 9 3 5.9 2.7 3.2 4.6 0.6 4.1 0.6 0.9S L PLA 13 11 3 6.7 5.7 1.0 5.5 4.2 1.3 4.2 0.3A S PLA 10 9 5 7.7 2.8 4.9 4.3 0.0 4.3 0.0 ª0.6B S PLA 12 8 3 8.1 4.1 4.1 4.4 1.6 2.8 1.6 ª1.2P S PLA 11 6 3 11.1 7.7 3.5 4.7 0.0 4.7 0.0 1.2H A e-PTFE 9 7 3 9.6 9.1 0.5 5.0 4.1 1.0 4.1 0.5P B e-PTFE 10 3 3 7.4 6.8 0.5 4.9 1.9 2.9 1.9 2.5N A e-PTFE 11 9 4 6.0 5.3 0.7 4.1 2.2 1.8 2.2 1.2Å J e-PTFE 9 6 6 9.6 8.7 1.0 5.2 3.8 1.4 3.8 0.4L-OK e-PTFE 10 5 3 8.7 6.4 2.3 3.5 0.0 3.5 0.0 1.1L M e-PTFE 7 4 4 5.8 3.5 2.3 3.2 1.0 2.2 1.0 ª0.1C-G M e-PTFE 10 8 3 10.6 5.2 5.4 6.4 1.6 4.9 1.6 ª0.5G P e-PTFE 10 6 4 9.2 5.6 3.6 6.4 0.0 6.4 0.0 2.8A N e-PTFE 10 7 4 8.2 6.3 1.9 3.9 1.6 2.4 1.6 0.5

Diff.Ωdifference; Crestal RAΩCrestal resorption or apposition

surgery (Table 5) showed a high corre-lation (RΩ0.77), the slope of the re-gression line being 0.88. In addition, thedifference between mean values for PBLat baseline and clinical ABL measure-ments during surgery (0.3∫1.1 mm)(p±0.05) was insignificant. The corre-lation between clinical measurements ofABL and IBD during surgery andradiographic ABL and IBD values(Table 5), on the other hand, was poor.Clinical measurements during surgeryfor ABL and IBD were significantlylarger than the corresponding radio-graphic values measured at baseline.The mean (SD) difference between clin-ical measurements and radiographicABL values at baseline was 1.7∫2.0mm (p∞0.01) and between clinical andradiographic IBD values at baseline1.9∫2.0 mm (p∞0.001). The width ofthe bone defects varied between 3 and6 mm.

Table 4 depicts mean (SD) values for

clinical measurements during surgeryand radiographic bone measurements.The clinical IBD depth averaged7.7∫1.6 mm at PLA-treated sites and6.1∫1.9 mm at e-PTFE treated sites.The corresponding radiographic IBDvalues measured 5.3∫0.9 mm and4.7∫1.2 mm respectively. There was nosignifiant difference in baseline data be-tween treatment groups (p±0.05).

Both treatments resulted in significantbone fill as evaluated by comparingbaseline and 12-month data for radio-graphic ABL (Table 4). Thus, bone fillaveraged 3.4∫1.2 mm following PLAtreatment (p∞0.001) and 2.0∫1.6 mmfollowing e-PTFE treatment (p∞0.01),the difference being statistically signifi-cant (p∞0.05). Together with a meancrestal resorption of 0.9 mm, a signifi-cant (p∞0.001) defect resolution wasachieved following both treatments aver-aging 4.4∫1.7 mm at PLA treated sitesand 3.0∫1.7 mm at e-PTFE-treated sites

338 Karapataki et al.

(p∞0.05). Both treatments also left a re-sidual intrabony defect of l.0∫1.4 and1.8∫1.4 mm respectively (p∞0.05).

Discussion

This study evaluated guided tissue re-generation (GTR) treatment of intra-bony defects distal to mandibular M2and compared resorbable and non-re-sorbable barriers. The defects were di-agnosed 5 years or more following sur-gical removal of the adjacent impactedM3. These defects are often referred toas residual bone defects of the previoussurgical procedure. It is possible,though, that the periodontal lesionswere already present at the distal aspectof the M2 at the time of the M3 re-moval. Undiagnosed, the lesion was nottreated at the time, and the diseasedroot affected wound healing in the areaand developed into a persistent intra-bony defect, to be diagnosed severalyears later and, possibly, at a more ad-vanced stage.

In recent years, a number of re-sorbable GTR barriers have beenintroduced. (For review see Gottlow etal. 1993, Greenstein & Caton 1993,Mendieta & Williams 1994). Resorbablebarriers may have a great advantageover non-resorbable barriers providedtheir efficacy is at least equal. First,only one surgical procedure is required,which saves time and money and issafer and less troubling for the patient.Second, unnecessary trauma to newly-formed tissue that would be caused bya second intervention and might nega-tively influence healing is avoided. Thepresent study as well as the recentstudies by Hugoson et al. (1995) onclass II furcation defects and by Roc-cuzzo et al. (1996) on buccal gingivalrecessions demonstrate that both thesafety and the efficacy of PLA re-sorbable barriers are at least equivalentto those of the commonly used non-re-sorbable e-PTFE barriers.

The results obtained in this studymust be evaluated against the accu-racy of the methods used for soft andhard tissue measurements. PDs andPALs were measured to the nearest 1mm using a manual, 1-mm gradedperiodontal probe. The reproducibilityand thereby the precision of the meas-urements made with manual probes,as compared to pressure sensitiveprobes, have been questioned. Recentstudies (Wang et al. 1995, Mayfield etal. 1996, Shilby et al. 1996), however,

have shown that the reproducibility ofPD and PAL measurements are equalto or even better with 1-mm gradedmanual probes than with pressure sen-sitive probes. In a recent study (Falket al. 1997), duplicate measurementsof 240 sites performed by the exam-iners in the present study, showed 95%agreement within ∫1 mm for PD andPAL measurements. Standarddeviations for differences between re-peated measurements of the 240 sitesaveraged 0.6 mm for intraindividualand 0.7 mm for interindividual com-parisons; the correlation betweenduplicate measurements was 0.92.Thus, in the present study, PD andPAL differences or changesΩ1.5 mm(Ω2¿SD) can be considered real.

Bone changes were evaluated bybone probing during local anaesthesiaand by linear measurements of radio-graphs. Initial PBL measurements didnot differ significantly from the clin-ical ABL measurements performedduring surgery, i.e., the distance fromthe same reference point on the toothto the bottom of the defect as meas-ured during surgery. Thus, the accu-racy of clinical bone probing measure-ments of bone height at baseline washigh. The threshold for changes inPBL levels would therefore be low andchanges or differencesΩ1 mm con-sidered real. It can also be assumedthat PBL at 12 months represents a‘‘re-entry’’ measurement.

Bone measurements on radiographsshowed a very high reproducibility. Onthe other hand, however, radiographssignificantly underestimated actualABL and IBD depths by nearly 2mm. Although these differences couldbe explained by distortion of theradiographic image of the tooth, theyare more likely a consequence of thebone anatomy in this area. While themandible widens considerably in theapical direction, the intrabony defectnarrows, rendering the apical portionof the defect undetectable in theradiograph. Similar discrepancies be-tween clinical and radiographic meas-urements of intrabony defects weredemonstrated (Tonetti et al. 1993, Ei-kholz et al. 1996, Falk et al. 1997). Itis therefore reasonable to suggest thatthe amount of bone fill measuredradiographically, although significant,was underestimated.

Having considered the measurementerrors, it can be concluded that bothtreatments resulted in statistically sig-

nificant PD reduction, PAL gain, andPBL and gain, as well as radiographi-cally detectable bone fill and defect res-olution. The results were clinically rel-evant. Others have found similar levelsof PAL gain and bone fill followingGTR treatment of intrabony defectsusing non-resorbable (Gottlow et al.1986, Becker et al. 1988, Becker &Becker 1993, Cortellini et al. 1993a,b,Cortellini et al. 1995) as well as re-sorbable barriers (Laurell et al. 1994,Cortellini et al. 1996, Falk et al. 1997).Great individual variations in treatmentoutcome were found. Although for theentire group of patients changes in PD,PBL and defect depth following treat-ment correlated significantly with thebaseline values for the respective vari-ables, the initial measures did not inter-act with the difference in outcomes be-tween treatments.

The difference in outcomes betweenthe two treatments may be related tothe early healing event. Inflammation inthe soft tissue covering the barriers oc-curred more frequently at sites treatedwith the non-resorbable e-PTFE bar-riers (Table 1). This inflammation couldhave been caused by micro-organismsaccumulating in the pouch formed be-tween the flap tissue and the barrier dueto lack of integration between the two.The resorbable PLA barrier used in thisstudy had a particular design promot-ing integration with the surroundingtissue along its entire surface to preventpouch formation and plaque accumu-lation. Accumulation of pathogens onthe submerged portion of e-PTFE bar-riers has been shown to be detrimentalto periodontal regeneration (Nowzari etel. 1995). Another reason may be thetrauma caused by the surgical removalof the barrier and perhaps the introduc-tion of infection at this time to the deli-cate, newly-formed tissue under thebarrier. Thus, Cortellini at al. (1993b)showed that about 25 % of the possibleregeneration found under the barrier atthe time of its removal had been lost atthe 1-year examination. Gain of clinicalattachment and bone 2 mm or morewas achieved in all patients except forone patient treated with the e-PTFEbarrier. This patient exhibited severeflap sloughing with early barrier ex-posure and did also have a poor plaquecontrol. If this ‘‘out-lyer’’ is excludedfrom the statistical analysis the differ-ences in treatment outcome still existbut do not reach the level of statisticalsignifcance. Therefore, within the limits

GTR treatment of intrabony defects 339

of this study, it can be concluded thatGTR treatment of intrabony defectsdistal to mandibular second molars willpredictably result in significant PD re-duction, PAL gain, and bone fill, andthat resorbable PLA barriers are atleast equivalent to non-resorbable e-PTFE barriers.

Acknowledgements

This study was supported by grantsfrom the Jonkoping County Council,Jonkoping, Sweden, and Guidor AB,Goteborg, Sweden.

Zusammenfassung

Die Heilung nach GTR-Behandlung von kno-chernen Defekten distal von 2. unteren Mola-ren unter Nutzung resorbierbarer und nicht re-sorbierbarer MembranenDie Ziele der vorliegenden randomisierten kli-nischen Studie waren (i) die Heilung von paro-dontalen intraalveolaren Defekten distal der2. unteren Molaren unter Nutzung resorbier-barer Polylactid-(PLA)-Membranen odernicht resorbierbarer Polytetrafluorethylen-(e-PTFE)-Membranen zu uberprufen und (ii)den therapeutischen Effekt der bioresorbier-baren versus der nicht resorbierbaren Mem-bran zu vergleichen. 19 Patienten mit intraal-veolaren Defekten distal der 2. unteren Mola-ren $4 mm (auf Rontgenbildern) wurden indie Studie einbezogen. Alle Defekte bestanden5 Jahre nach der chirurgischen Entfernungvon impaktierten 3. Molaren. Nach der Lap-penmobilisation und der Defektreinigungwurden die Defekte randomisiert, bedecktentweder mit einer resorbierbaren PLA odermit einer nicht resorbierbaren e-PTFE-Mem-bran. Die Lappen wurden reponiert und ver-naht, um die Membran vollstandig zu bedek-ken. Die Behandlung wurde 1 Jahr spater kli-nisch evaluiert durch Messung derSondierungstiefe (PD), des Attachmentnive-aus (PAL) und des Knochenniveaus (PBL) so-wie rontgenologisch durch Messung des Kno-chenniveaus an computerdigitalisierten Bil-dern von Rontgenaufnahmen, die vor derTherapie und ein Jahr postoperativ genom-men wurden. Beide Therapiemodalitatenfuhrten zu einer signifikanten PD-Reduktion,PAL-Gewinn und Knochenfullung. Die insge-samte PD-Reduktion war 5.3∫1.9 mm bei denPAL-behandelten Flachen und 3.7∫1.7 mmbei den e-PTFE-behandelten Flachen(p∞0.05). Die korrespondierenden Werte furPAL-Gewinn waren 4.7∫0.7 mm und 3.6∫1.7mm (p∞0.05) und fur PBL-Gewinn 5.1∫1.2und 3.3∫2.0 mm (p∞0.05). Die durchschnitt-liche rontgenologisch festgestellte Knochen-fullung betrug 3.4∫1.2 fur die PAL- und2.0∫1.6 mm fur die e-PTFE-Membranflachen(p∞0.05). Die Ergebnisse der rontgenologisch

durchgefuhrten Knochenniveaumessungenwaren signifikant kleiner als die entsprechen-den klinischen Messungen, was zeigt, daß dieRontgenbilder eine Tendenz zur Unterbewer-tung der Knochenfullung aufweisen. Zusam-menfassend erbringt die Behandlung tiefer in-traalveolarer Defekte distal der 2. unterenMolaren unter Nutzung von resorbierbarenPLA-Membranen eine signifikante Reduktionder Sondierungstiefe, PAL-Gewinn und Kno-chenfullung mindestens so erfolgreich wie diesbei nicht resorbierbaren e-PTFE-Membranenist.

Resume

Guerison suite au traitement GTR de lesionsintraosseuses en distal des 2emes molairesmandibulaires en utilisant des barrieres resor-bables et non-resorbablesLes objectifs de l’etude clinique presenteetaient (i) d’evaluer la guerison de lesions in-traosseuses parodontales en distale desdeuxiemes molaires mandibulaires en utili-sant des membranes en acide polylactique re-sorbables (PLA) et non-resorbables (e-PTFE) en teflon, (ii) de comparer l’effet the-rapeutique des barrieres resorbables VS non-resorbables. Dix-neuf patients presentant deslesions intraosseuses en distale des secondesmolaires >4 mm jugees sur base de radio-graphies ont ete inclus dans cette etude. Leslesions etaient toujours presentes cinq anneesapres extraction chirurgicale des dents de sa-gesse. Apres elevation du lambeau et nettoya-ge de la lesion, celle-ci a ete ensuite recouvertau hasard soit d’une membrane PLA soitd’une e-PTFE. Les lambeaux ont ete reposi-tionnes et sutures pour recouvrir complete-ment cette derniere. Le traitement a ete eva-lue cliniquement apres un an en mesurant laprofondeur au sondage (PD), le niveau d’at-tache au sondage (PAL) et le niveau osseuxau sondage (PBL) et radiographiquementpar les mesures des niveaux osseux sur desimages digitalisees de radiographies, prisesimmediatement avant la chirurgie et un anapres. Les deux types de traitement se sontaccompagnes d’une reduction significative dePD, d’un gain de PAL et d’un gain osseux.La reduction totale de PD etait de 5.3∫1.9mm pour les sites PLA et de 3.7∫1.7 mmpour les sites e-PTFE (p∞0.05). Les valeurscorrespondantes pour le gain de PAL etaient4.7∫0.7 mm et 3.6∫1.7 mm (p∞0.05) et pourle gain de PBL 5.1∫1.2 mm et 3.3∫2.0 mm(p∞0.05). Le comblement osseux radio-graphique etait de 3.4∫1.2 mm pour les PLAet de 2.0∫1.6 mm pour les e-PTFE (p∞0.05).Les mesures du niveau osseux radiographi-que etaient significativement plus petites queles mesures cliniques correspondantes, indi-quant que les radiographies tendent a sous-estimer le comblement osseux. En conclu-sion, le traitement GTR de lesions intraos-seuses profondes en distale des deuxiemesmolaires inferieures par des membranes PLAresorbables s’est accompagne d’une reduc-tion significative de PD, d’un gain de PAL et

d’un comblement osseux au moins equivalenta ce qui etait obtenu avec les membranes bar-rieres e-PTFE non-resorbables.

References

Axelsson, P. & Lindhe, J. (1974) The effect ofa preventive programme on dental plaque,gingivitis and caries in schoolchildren. Re-sults after one and two years. Journal ofClinical Periodontology 1, 126–138.

Becker, W., Becker, B., Berg, L., Prichard, J.,Caffesse, R. & Rosenberg, E. (1988) Newattachment after treatment with root iso-lation procedures: Report for treated ClassIII and Class II furcations and verticalosseous defects. International Journal ofPeriodontics and Restorative Dentistry 3,9–23.

Becker, W. & Becker, B. (1993). Treatment ofmandibular 3-wall infrabony defects byflap debridement and expanded polyte-traflouroethylene barrier membranes.Long-term evaluation of 32 treated pa-tients. Journal of Periodontology 64, 1138–1144.

Blomlof, L., Lindskog, S., Appelgren, R.,Jonsson, B., Weintraub, A. & Hammar-strom, L. (1987) New attachment in mon-keys with experimental periodontitis withand without removal of the cementum.Journal of Clinical Periodontology 14, 136–143.

Chin Quee, T., Gosselin, D., Millar, E. &Stamm, J. (1985) Surgical removal of fullyimpacted mandibular third molar. The in-fluence of flap design and alveolar boneheight on the periodontal status of the sec-ond molar. Journal of Periodontology 56,625–630.

Cortellini, P., Pini Prato, G. & Tonetti, M.(1993a) Periodontal regeneration of hu-man infrabony defects (I). Clinical meas-ures. Journal of Periodontology 64, 254–260.

Cortellini, P., Pini Prato, G. & Tonetti, M.(1993b) Periodontal regeneration of hu-man infrabony defect (II). Re-entry pro-cedures and bone measures. Journal ofPeriodontology 64, 261–268.

Cortellini, P., Pini Prato, G. & Tonetti M.(1995) Periodontal regeneration of humanintrabony defects with titanium-reinforcedmembranes. A controlled clinical trial.Journal of Periodontology 66, 797–803.

Cortellini, P., Pini Prato, G. & Tonetti, M.(1996) Periodontal regeneration of humanintrabony defects with bioresorbable mem-branes. A controlled clinical trial. Journalof Periodontology 67, 227–223.

Demolon, I., Persson, G., Moncla B., John-son, R. & Ammons, W. (1993) Effects ofantibiotic treatment on clinical conditionsand bacterial growth with guided tissue re-generation. Journal of Periodontology 64,609–616.

Eikholz, P., Benn, D. & Staehle, H. (1996)Radiographic evaluation of bone re-generation following periodontal surgerywith and without expanded polytetraflou-

340 Karapataki et al.

roethylene barriers. Journal of Periodonto-logy 67, 379–385.

Falk, H., Laurell, L., Ravald, N., Teiwik,A. &. Persson, R. (1997) GTR therapy of203 consecutively treated infrabony defectsusing a bioresorbable matrix barrier. Clin-ical and radiographic findings. Journal ofPeriodontology 68, 571–581.

Gottlow, J., Nyman, S., Lindhe, J., Karring,T. & Wennstrom, J. (1986) New attach-ment formation in the human periodon-tium by guided tissue regeneration. Casereports. Journal of Clinical Periodontology13, 604–616.

Gottlow, J., Nyman, S. & Karring, T. (1992)Maintenance of new attachment gainedthrough guided tissue regeneration.Journal of Clinical Periodontology 19, 315–317.

Gottlow, J. (1993) Guided tissue regenerationusing bioresobable and non-resorbable de-vices: Initial healing and long-term results.Journal of Clinical Periodontology 64,1156–1165.

Greenstein, G. & Caton, J. (1993) Biodegrad-able barriers and guided tissue regenera-tion. Periodontology 2000 1, 36–45.

Grondahl, H. G. & Lekholm, U. (1973) In-fluence of mandibular third molars on re-lated supporting tissues. InternationalJournal of Oral Surgery 2, 137–142.

Hugoson, A. & Kugelberg, C. F. (1987) Theprevalence of third molars in a Swedishpopulation. An epidemiological study.Community Dental Health 5, 121–138.

Hugoson, A., Ravald, N., Teiwik, A., For-nell, J., Johard, G. & Gottlow, J. (1995)Treatment of furcation class II involve-mentis in humans with bioresorbable andnonresorbable GTR barriers. Journal ofPeriodontology 66, 624–634.

Jeffcoat, M., Williams, R. & Ruttiman, R.(1984) Relationship between linear andarea measurements of radiographic bonelevels utilising simple computerised tech-

niques. Journal of Periodontal Research19, 191–198.

Karapataki, S., Hugoson, A. & Kugelberg,C. F. (2000) Healing following GTR treat-ment of bone defects distal to mandibularsecond molars after surgical removal ofimpacted third molars. J. Clin. Peri-odontol. 27, in press.

Kugelberg, C. F., Ahlstrom, U., Ericson, S.,Hugoson, A. (1985) Periodontal healingafter impacted lower third molar surgery.International Journal of Oral Surgery 14,29–40.

Kugelberg, C. F. (1990) Periodontal healing2 and 4 years after impacted lower thirdmolar surgery. A comparative retrospec-tive study. International Journal of Oraland Maxillofacial Surgery 19, 341–345.

Laurell, L., Falk, H., Fornell, J., Johard,G. & Gottlow, J. (1994) Clinical use of abioresorbable matrix barrier in guidedtissue regeneration therapy. Case series.Journal of Periodontology 65, 967–975.

Marmary, Y., Brayer, L., Tzukert, A. & Fell-er, L. (1985) Alveolar bone repair follow-ing extraction of impacted mandibularthird molars. Oral Surgery, Oral Medicineand Oral Pathology 60, 324–326.

Mayfield, L., Bratthall, G. & Attstrom R.(1996) Periodontal probe precision using 4different periodontal probes. Journal ofClinical Periodontology 23, 76–82.

Mendieta, C., & Williams R. (1994) Peri-odontal regeneration with bioresorbablemembranes. Current Opinion in Periodon-tology 4, 157–167.

Nowzari, H., Matian F. & Slots, J. (1995)Periodontal pathogenes on polytetraflou-roethylene membrane for guided tissue re-generation. Journal of Clinical Periodonto-logy 22, 469–474.

Pecora, G., Celetti, R., Davarpanah, M., Co-vani, U. & Etienne, D. (1993) The effect ofguided tissue regeneration on healing afterimpacted third molar surgery: 1-year re-

sults. International Journal of Periodonticsand Restorative Dentistry 13, 397–407.

Roccuzzo, M., Lungo, M., Corrente, G. &Gandalfo, S. (1996) Comparative study ofa bioresorbable and a non-resorbablemembrane in the treatment of human buc-cal gingival recessions. Journal of Period-ontology 67, 7–14.

Shilby, O., Nashef, N., Andreana, S., Cian-cio, S., Kharchaf, M. & Tabbaa, S. ( 1996)Comparison of manual and automatedprobing of attachment loss. Journal ofDental Research 75, (special issue), abstr.1541.

Silness, J. & Loe, H. (1964) Periodontal dis-ease in pregnancy (II). Correlation be-tween oral hygiene and periodontal con-ditions. Acta Odontologica Scandinavica22, 121–135.

Stanely, H., Alattar, M., Collet, W., Stringfel-low, J. & Spiegel, E. (1988) Pathologicalsequelae of ‘‘neglected’’ impacted thirdmolars. Journal of Oral Pathology 17, 113–117.

Tonetti, M., Pini-Prato, G., Williams, R. &Cortellini, P. (1993) Periodontal regenera-tion of human infrabony defects (III). Di-agnostic strategies to detect bone gain.Journal of Periodontology 64, 269–277.

Wang, S-F., Leknes, K., Zimmerman, G., Si-gurdsson, T., Wikesjo, U. & Selvig, K.(1995) Reproducibility of periodontalprobing using a conventional manual andan automated force-controlled electronicprobe. Journal of Periodontology 66, 38–46.

Address:

Anders HugosonPublic Dental Service AdministrationP O Box 1024SE-551 11 JonkopingSweden