Studies to date have reached differing conclusions regarding the long-term prognosis of teeth with class III furcation involvement. Replacement of such teeth with implants could be an alternative. This report compares the treatment outcomes of 2 cases with similar disease progression: 1 treated by implant therapy and 1 maintained with nonsurgical periodontal treatment. Two patients with advanced chronic periodontitis and class III furcation involvement of all molars were treated. Case 1 received a conservative periodontal and antibiotic treatment, followed by 15 years of maintenance. In case 2, the molars were extracted and replaced with implants, and the implants were observed for 7 years. Clinical attachment level (CAL), probing attachment level (PAL), bleeding on probing, plaque index, and periodontal pathogens were recorded. Despite good compliance of case 1, periodontal pathogens were not eliminated and tissue destruction was not halted. The PAL outcomes of case 2 improved over time; mean PAL loss reached 0.35 mm/y in the first 3 years and then decreased to 0.01 mm/y. While CAL outcomes did not change in case 2, case 1 showed increased CAL loss after 8 years. Based on the limited findings of this case report, extraction of molars with class III furcation involvement and subsequent implant placement may render a better predictability of treatment outcomes than nonsurgical periodontal therapy in the cases of infection with periodontal pathogens.
The main goal of nonsurgical periodontal therapy is to change the subgingival environment by re-establishing a microbial flora compatible with health. Depending on disease progression, this may be achieved through mechanical instrumentation with or without the systemic use of antibiotics. Therapy reduces the total bacterial load and shifts the microflora from a strictly anaerobic to a facultative anaerobic microbiota.1,2
Supportive periodontal treatment (SPT) aims to maintain periodontal health and prevent disease recurrence and further attachment loss.3 This method controls supragingival and subgingival plaque by reducing the total bacterial load and attempting to keep pathogenic species below detectable levels.4,5 Although maintenance is effective for most periodontitis patients, individuals suffering from advanced periodontal disease with furcation involvement of multirooted teeth may lose teeth or exhibit periodontal breakdown despite regular SPT.1 Patients with advanced periodontal destruction and furcation involvement are highly susceptible to disease and may be at increased risk of developing recurrent periodontal breakdown following therapy. In these patients, horizontal and vertical periodontal attachment loss may develop in the affected areas. The poorer prognosis of furcation-involved teeth is confirmed by longitudinal studies analyzing the outcomes of periodontal therapy. Huynh-Ba et al6 discussed the data of Hirschfeld and Wassermann,7 showing that patients following SPT lost 31% of multirooted teeth with furcation involvement over 22 years.
For these reasons, multirooted teeth represent a challenge for the clinician. Root resective therapy is a surgical option for molars with class III furcation involvement.8 Not all patients consent to surgical treatment or molar extraction, however. In these cases, the clinician may attempt to salvage such teeth with nonsurgical periodontal therapy, antimicrobial treatment, and SPT.
Fully or partially edentulous dentitions with dental implants frequently show reliable long-term prognosis.9–12 A recent retrospective study compared the complication and survival rates of root-resected mandibular molars with those of dental implants replacing mandibular molars.8 The authors found a greater overall risk of complications for resected molars than for implants over a 4-year maintenance period (odds ratio = 3.79). The incidence risk for nonsalvageable complications (failures) was almost 10 times greater for resected molars than for implants (odds ratio = 10.11).
This study compared clinical and microbiological statuses in the dentitions of 2 patients with advanced chronic periodontitis including molars with class III furcation involvement. These patients followed different treatment protocols. Case 1 completed initial periodontal treatment followed by SPT, and case 2 followed a treatment protocol in which the target teeth were extracted and replaced by dental implants. Follow-up periods were 15 years for case 1 and 7 years for case 2.
Materials and Methods
The 2 patients were referred to the private practice of one of the authors (G.-G.K.Z., Duesseldorf, Germany) for treatment. Case 1 was a 33-year-old male (presented in March 1993), and case 2 was a 39-year-old female (presented in February 2000). Both patients were nonsmokers, were in good general health, and were not taking any medications or drugs. They reported more than 8 years of periodontal problems, including increased periodontal probing depths and bleeding on probing (BOP), tooth extractions due to advanced bone loss and/or tooth mobility, and repeated surgical periodontal treatment. The surgical treatments were not accompanied by microbiological assessment and were performed by different surgeons. It was not possible to reconstruct the case history or obtain the patient's records in either case. Both patients relocated and therefore changed their dentists 12 months before presenting in our office. During these 12 months, they were treated and placed on a 3-month maintenance schedule at the same dental office. Maintenance visits consisted of supragingival and subgingival debridement, tooth polishing, and oral hygiene instruction.
At the initial examination, both patients presented with no pain, good oral hygiene (plaque index [PI13] 10% in case 1 and 8% in case 2; BOP 5% in both cases), and no signs of acute inflammation. Maxillary and mandibular molars showed class III furcation involvement. Furcations of the mandibular molars could be buccally and lingually probed, while those of the maxillary molars could be buccally, palatal-mesially, and palatal-distally probed. Radiological evaluation revealed generalized advanced bone loss extending ≥50% of the root length, as well as furcation involvement. Both cases were diagnosed as chronic periodontitis with advanced loss of periodontal support.14
Patients received detailed information on treatment options and risks (described below). They were given at least 1 week to consent to treatment, and the treatments were performed in accord with the Helsinki Declaration of 1975, as revised in 1983.
BOP, probing attachment level (PAL), and clinical attachment level (CAL) were measured (to the nearest half-millimeter) using the same periodontal probe (UNC15, Hu-Friedy, Leimen, Germany). The PAL and CAL were recorded on the 4 surfaces of the implants and natural teeth, respectively. The PAL was defined as the distance (mm) between the deepest point of the peri-implant pocket and the smooth neck section of the implants; CAL was defined as the distance (mm) between the deepest point of the pocket and the cemento-enamel junction or the margin of the fixed partial denture (FPD). Plaque accumulation was measured with a plaque relevator (Mira-2-Ton, Hager & Werker, Duisburg, Germany) and recorded as presence or absence of plaque (±).
The patient had undergone surgical periodontal treatment twice during the previous 3 years (in 1990 and 1992) to address recurrent periodontal problems. Teeth #5, 12, 21, and 28 had already been extracted (Figures 1 and 2).
The mean CAL was 7.5 mm (range, 5–12 mm) (Table 1). At the first examination (E1), subgingival plaque samples were collected with sterile paper points from the deepest pocket of each quadrant. The samples were sent to a commercial laboratory, which detected Tannerella forsythensis (Tf; previously Bacteroides forsythus; 5 × 104), Campylobacter rectus (Cr; 5 × 103), Fusiformis nucleatum (Fn; 5 × 103), Porphyromonas gingivalis (Pg; 5 × 103), and Treponema denticola (Td; 1 × 104). Oral hygiene instructions were given, and scaling and root planing (SRP) with supragingival debridement and tooth polishing was performed. Metronidazole (MTNZ; 800 mg/d) was prescribed for 1 week following SRP (Table 1). Ten weeks after the combined mechanical and antibiotic treatment (E2), plaque samples were taken and no periodontal pathogens were detected.
Extraction of the molars and placement of either conventional or implant-supported restorations was recommended. The patient refused extraction and opted for a semiannual maintenance schedule (described below), beginning at E2.
Periodontal pathogens were detected in this patient's subgingival plaque samples in 2004 (E3: Pg 14.72 × 106; Tf 10.17 × 106; Td 3.42 × 106) and 2008 (E4: Pg 1.66 × 106; Tf 6.97 × 106) (Table 1).
The patient had undergone surgical periodontal treatment 3 times during the previous 5 years (in 1995, 1997, and 1999) to address recurrent periodontal problems. Teeth #1, 2, and 30–32 had been extracted. The mean CAL was 8.5 mm. All teeth had class I to II mobility and several vertical defects extending toward the apical third of the teeth (Figure 3). Subgingival plaque samples were taken with sterile paper points from the deepest pocket of each quadrant and sent to a commercial microbiological laboratory. Analysis detected the pathogens Tf (12 × 106), Pg (15 × 106), and Td (5 × 106).
All maxillary teeth (#3–16) and the mandibular teeth (#17–19, 23–26, and 29) were extracted, and the extraction sockets were augmented as previously described.16 Briefly, an intrasulcular incision extending to the adjacent teeth was made, and a full-thickness flap was elevated. No vertical releasing incisions were made. Teeth were extracted, and the sockets were carefully curetted and irrigated with sterile saline solution. Extraction sockets were covered with nonresorbable dPTFE membrane (Cytoplast, Regentex GBR-200, Osteogenics Biomedical, Lubbock, Tex) without the use of any soft- or hard-tissue grafts. No further steps were taken to secure the membrane in place, and the flap was repositioned and secured with interrupted sutures. The membranes were left partially exposed during the healing period.15 Area #30 received a crestal onlay bone graft harvested from the mandibular retromolar region. The block graft was fixated with titanium screws (Memfix screw Ø 1.2 mm × 4.5 mm, Straumann, Waldenburg, Switzerland) and covered with a resorbable membrane (BioGide, Geistlich Biomaterials, Wolhusen, Switzerland). The remaining natural teeth (#20–22, 27, and 28) were treated by access flap surgery. An interim full denture for the maxilla and a removable partial denture for the mandible were delivered 1 day after surgery.
Six months after extractions and augmentations, and after consultation with an ear, nose, and throat specialist to rule out sinus pathologies, bilateral sinus augmentation and implant placement were performed. Sinus augmentation was performed according to the procedure described by Beaumont et al.16 The graft was a 1∶1 mixture of xenograft (BioOss spongiosa 0.25–1 mm, Geistlich Biomaterials, Wolhusen, Switzerland) and autogenous corticocancellous bone (harvested from the retromolar pad or chin area). The access window was closed using an absorbable barrier (BioGide, Geistlich Biomaterials, Wolhusen, Switzerland).
Twelve dental implants were placed (maxilla, n = 7: #3, 6, 8, 9, 11, 12, 14; mandible, n = 5: #19, 23, 24, 26, 30). Cylindrical screw-type implants (RN, screw cylinder SLA, Straumann, Waldenburg, Switzerland) were placed using a 2-stage surgical approach. Following full-thickness flap elevation, osteotomy site preparation was performed at 875 rpm, and implants were manually placed at a torque of 35 Ncm.
Four months after placement, the implants were loaded. The maxillary dentition was restored with a telescopic-crown–retained palate-free removable denture, and the mandible was restored with FPDs (Figure 4).
Medication and postoperative care
Systemic medications were prescribed when periodontal pathogens were detected. For initial infection by Pg alone or with other pathogens (except Haemopuilus actinomycetemcomitans), 800 mg/d MTNZ was prescribed for 1 week17 and a 0.1% chlorhexidinedigluconate rinse (Chlorhexamed Fluid, GlaxoSmithKline, Buehl, Germany) was prescribed twice daily. The patients were instructed to start these medications immediately after SRP (case 1) or 1 day before extractions and implant or periodontal surgery (case 2).
In case 2, surgical sutures were left in for 1 week, and the dPTFE membranes were removed after a healing period of 4 weeks. In the period between surgery and implant loading, the patient received oral hygiene instructions, and debridement and polishing were performed once a month. Postoperative management following sinus augmentation included a systemic antibiotic (500 mg amoxicillin, twice daily for 7 days), an anti-inflammatory nasal spray (mometasone furoate monohydrate, twice daily for 7 days), and an analgesic. The patient was instructed to avoid using any removable appliance for the first 2 postoperative weeks.
Both patients were assigned to a semi-annual maintenance schedule. During maintenance appointments, oral hygiene instructions were given, supragingival deposits were removed, restored teeth and implants were polished, and subgingival irrigation with 0.1% chlorhexidine digluconate solution (Chlorhexamed Fluid, GlaxoSmithKline, Buehl, Germany) was performed. Subgingival plaque samples were taken annually. When periodontal pathogens were detected, full-mouth SRP was performed in 1 session and adjunctive systemic antibiotic therapy was administered. Plaque samples were repeated 10 weeks after this treatment.
Regarding case 2, the examination at 3 months after FPD placement was considered as the baseline (BL), and the patient was enrolled in a semi-annual maintenance program.
The CAL outcomes at each examination are presented as the mean ± SD and range. The BOP and PI are expressed as a percentage of affected sites. From E2 to E4, tooth sites were grouped according to CAL changes (gain, loss, or no change) in comparison to both E1 and the previous examination. The statistical units were the teeth of only 1 patient, preventing the use of many traditionally employed statistical analyses. Cohen's d effect sizes (ES) were calculated to validate observed CAL gains and to estimate differences in CAL outcomes between the maxilla and mandible at E1, E3, and E4 as well as maxilla, mandible, and total differences between E1 and E4, E1 and E3, and E3 and E4. Cohen's d was also calculated to estimate CAL differences between maxillary and mandibular single- and multirooted teeth at E1, E3, and E4. The CAL changes for these groups of teeth between E1 and E3, E1 and E4, and E3 and E4 were also evaluated. The ES was defined as small, d = 0.2; medium, d = 0.5; or large, d = 0.8.
Changes in CAL and/or PAL outcomes were evaluated using an analysis of variance (ANOVA) with 1 repeated measure (evaluation time). The mean CAL was calculated using data from all tooth surfaces (distal, lingual, buccal, and mesial). Evaluations were conducted at BL and at 1 (E1), 3 (E3), 5 (E5) and 7 (E7) years thereafter. Follow-up repeated-measure ANOVA was performed to determine whether changes in CAL or PAL differed by jaw (maxilla vs mandible), region (anterior vs posterior), or tooth surface (distal, lingual, buccal, or mesial). The CAL and PAL were separately evaluated for all statistical analyses. The PI and BOP were considered dichotomous variables, classified as either present or absent. These variables were descriptively analyzed and are presented as percentages of affected sites at each evaluation. Fisher exact test was used to determine whether the presence of plaque or bleeding differed by type (implant vs natural tooth), jaw (maxilla vs mandible), or region (anterior vs posterior).
No adverse reactions to prescribed medication or local anesthesia were observed during the study period of 15 years. The treatment performed from E1 to E4 did not eliminate all periodontal pathogens; reinfections with Pg and Tf occurred 11 and 15 years after the initial examination (E3 and E4) (Tables 1 and 2).
No significant changes in total CAL outcomes were observed during the first 11 years of the study period (E1 to E3) (Tables 1 and 2). Total CAL outcomes increased (attachment loss) between E3 and E4 (Table 1). Maxillary teeth of all types showed higher CAL outcomes over time than mandibular teeth (Table 3). Maxillary and mandibular CAL outcomes were higher in multirooted teeth than in single-rooted teeth (Table 3).
The maxillary teeth showed a mean CAL loss of 1.2 mm between E1 and E2 (maxillary total CAL loss mean = 1.2 mm, median = 1 mm; maxillary molars CAL loss mean = 1.6 mm, median = 2 mm). Maxillary CAL was stable for the following 6 years (E2 to E3). A further mean CAL loss of 5.2 mm occurred between E3 and E4 (Table 3; Figure 1a and b). A mean CAL loss of 1 mm was observed for the maxillary incisors, canines, and premolars between E1 and E2, followed by no change until E4 (Table 3; Figure 1a and b). Multirooted maxillary teeth showed a mean CAL loss of 5.2 mm at E3 (Table 3; Figure 1a and b). CAL outcomes for all mandibular teeth were constant for 11 years (E1 to E3). A mean loss of 3.3 mm was observed between E3 and E4 (Tables 1 and 3; Figure 1c and d). During the last 4 years of observation, mandibular molars and all multirooted teeth lost a mean of 7.3 mm, while mandibular incisors, canines, and all single-rooted teeth showed CAL losses of only 0.3–0.4 mm (Table 3; Figure 1c and d).
Comparison of E1 and E3 CAL outcomes demonstrated a medium positive ES (d = 0.3) for all teeth, a large positive ES (d = 0.7) for maxillary teeth, and a small positive ES (d = 0.2) for mandibular teeth. Comparison of E1 and E4 yielded medium to large negative ES values (all teeth: large ES, d = −0.7; maxilla: medium ES, d = −0.4; mandible: large ES, d = −1.1). Large negative ES values were also found between E3 and E4 (all teeth: d = −1.0; maxilla: d = −0.9; mandible: d = −1.1) (Figure 5).
Comparison of CAL outcomes for maxillary single-rooted teeth revealed a medium positive ES (d = 0.4) between E1 and E3, a small positive ES (d = 0.2) between E1 and E4, and a small negative ES (d = −0.2) between E3 and E4. Maxillary multirooted teeth demonstrated a large positive ES (d = 0.9) between E1 and E3, a large negative ES (d = −1.7) between E1 and E4, and a medium negative ES (d = −0.4) between E3 and E4. Mandibular single-rooted teeth showed a small positive ES (d = 0.2) between E1 and E3, a small negative ES (d = −0.1) between E1 and E4, and a medium negative ES (d = −0.4) between E3 and E4. Mandibular multirooted teeth demonstrated a small positive ES (d = 0.1) between E1 and E4 and a large negative ES between E1 and E4 and between E3 and E4 (d = −4.6 and d = −5.1, respectively). Comparison of CAL outcomes between all single- and multirooted teeth showed negative ES in all cases (maxillary single- vs multirooted teeth: d = −0.7 at E1, d = −0.5 at E3, d = −2.9 at E4; mandibular single- vs multirooted teeth: d = −0.2 at E1, d = −0.3 at E3, d = −6.0 at E4) (Figure 5).
Because the data for this case were not normally distributed, Greenhouse-Geisser corrections were used for all statistical analyses. No significant changes were noted for CAL outcomes when the data were grouped by any of the variables (Table 4 ).
When averaged across all tooth surfaces, significant PAL loss was observed over time (P < .001); however, there were no noticeable trends in CAL outcomes over time (P = .53) (Table 4 ). Post hoc analysis revealed PAL loss until E3 (mean loss BL to E3 = 1.05 mm; 0.35 mm/y), after which no clinically meaningful or statistically significant changes were noted (mean loss E3 to E5 = 0.04 mm; 0.01 mm/y). When grouped by implant surface, the PAL data showed a significant time × group interaction effect (P = .033), suggesting that PAL changes were dependent on implant region. Although all regions demonstrated a PAL loss, the buccal region showed the largest loss, and the distal and lingual regions displayed a slightly smaller change. PAL time × group interaction effects were not significant when grouped by location (P = .127) but approached significance when grouped by jaw (P = .05), with the mandible demonstrating a slightly larger PAL loss than the maxilla (Figures 4 and 6).
Baseline BOP was higher than any subsequent evaluations, with 16 of 17 sites bleeding at BL and 11 to 12 of 17 sites bleeding at E1, E3, E5, and E7. BOP did not differ by type (implant vs natural), jaw (maxilla vs mandible), or region (anterior vs posterior) at any time (Table 4 ).
The PI showed no noticeable trend over time. The BL PI values were highest, affecting 14 of 17 sites. At E1, E3, E5, and E7, plaque was present on at least 1 surface of 11 to 13 sites. The PI did not differ based on type (implant vs natural), jaw (maxilla vs mandible), or region (anterior vs posterior) at any time. At E1, the difference approached significance (P = .053), with plaque present on at least 1 surface of more implants (11 of 12) than natural teeth (2 of 5) (Table 4 ).
Periodontal therapy and maintenance seek to preserve a healthy dentition through elimination of etiological factors, limit subgingival periodontal flora, and/or suppress periodontal pathogens and improve personal oral hygiene.18,19 Given the poor long-term prognosis of teeth with class III furcation involvement, it is debatable whether these teeth should be maintained or extracted and replaced by implants or other restorations.
Previous studies have established that multirooted teeth are at higher risk for periodontal breakdown and loss than single-rooted teeth and that the presence of furcation involvement leads to a higher incidence of periodontal breakdown in multirooted teeth.7,20
The clinical outcomes of case 1 showed high long-term mortality of both maxillary and mandibular molars with class III furcation involvement, after combined antibiotic and SRP treatment. This could be explained by various factors that contribute to high mortality of maxillary posterior teeth, such as difficulty of plaque removal, anatomic design of the roots, occlusal stresses, lack of distal bone support, problems of root proximity, and iatrogenic problems. A successful treatment protocol should minimize as many of these factors as feasible.20–23 Pg was detected in furcation areas at E3 and E4, despite mechanical/antimicrobial therapy and SPT. This finding is consistent with Flemming et al,24 who reported low predictability of antimicrobial therapy for suppression of Pg to undetectable levels. The greater CAL loss in maxillary (vs mandibular) molars may be explained by anatomical differences and the greater difficulty of cleaning the maxillary furcations. These factors may influence long-term prognosis of the teeth by favoring plaque accumulation.25 Plaque retention in furcation areas could also explain increased BOP levels, although PI outcomes (recorded only on tooth sites) did not change and pathogens decreased over time.
The results of this study agree with those previously published, although some studies did not differentiate between class I, II, or III furcation involvement. McFall20 reported the loss of 56.9% and Hirschfeld and Wassermann7 the loss of 31.4% of furcation-involved teeth with questionable prognosis during maintenance periods of 14 and 22 years, respectively. The lowest incidence of tooth loss occurred in the well-maintained groups (27.3%20 and 19.3%7), and the number of molars lost was higher in groups with poorer hygiene and/or less intensive participation in SPT programs (downhill group = 68.9%20 and 64.7%7, extreme downhill group = 92.3%20 and 88.4%7).
A comparatively low mortality of furcated molars has been reported in other studies.26–28 Carnevale et al26 found a 4.9% loss in furcation-involved molars after 3 to 6 years and a 1.6% loss after 7 to 11 years. Ramfjord et al27 reported that molars and maxillary bicuspids had a poorer prognosis than other teeth, and Ross and Thompson28 reported the loss of only 12% of 341 teeth with furcation involvement. This variation in tooth mortality may affect prognosis and treatment planning.7 However, since these studies did not differentiate between class II and class III furcation involvement, direct comparison to our results is not possible.
In case 1 of the present study, all multirooted teeth showed class III furcation involvement. These teeth were nonsurgically treated and survived for a period of 8 years. Hirschfeld and Wassermann7 studied a comparable group (their well-maintained group) and found that 80.7% of furcated teeth survived over 22 years. Surgical treatment and/or more frequent SPT may more effectively salvage these teeth.
Case 2 demonstrated an implant survival rate of 100% over 7 years. These results are consistent with those previously published.29–33 Using the same implant type as in the present report, Levine et al29 reported a cumulative survival rate of 99.1% for 675 posterior implants after 21.30 months of maintenance. A 10-year prospective study of 112 implants found survival rates of 90.5% in patients with a history of chronic periodontitis and 96.5% in patients without such history.30 Additional studies31–33 have demonstrated survival rates of 90.2% to 95.5% for maxillary implants and 88% to 95.2% for mandibular implants. A cumulative survival rate of 97% for molar implants over 15 years was demonstrated by Fugazzotto.34
Nonresorbable dPTFE membranes were used to regenerate the extraction sockets and prepare the areas for implant placement. Histological findings 8 months after surgery showed formation of new vital bone.8 One limitation of this study was the probing procedure, in which measurements were rounded up to the nearest half-millimeter. Despite this, PAL loss in the first 3 years reached 0.35 mm per year and thereafter declined to 0.01 mm per year. These findings are similar to those observed after implant placement and loading in native bone.
The higher PAL loss we observed at buccal sites (vs other sites) was due to recession of the soft tissue. This may be associated with the implant design (tulip) and/or because we did not perform soft- or hard-tissue augmentation in these areas. Although radiographs at E7 showed no bone loss, the private office in which this study was performed did not use standardized radiographs and thus may have limited observational capabilities.
Both cases presented here showed equal loss of supporting periodontal tissues in the first 8 years. Periodontal conditions in furcated areas deteriorated after this time, with clear clinical differences between implants and nonsurgical treatment. Nonsurgical periodontal therapy combined with biannual SPT was not sufficient to prevent further attachment loss around teeth with class III furcation involvement. Infection of the furcations by periodontal pathogens could not be efficiently controlled. The limited findings of this case report suggest that extraction of molars with class III furcation involvement and subsequent implant placement may render a better predictability of treatment outcomes than nonsurgical periodontal therapy.
The authors thank the DENTEGRIS Deutschland Implant Co, Germany, for statistical analysis of the data. Except for this support, DENTEGRIS Deutschland Implant Co did not provide any support, financial or material, to the authors or the patients reported herein. The authors report no conflicts of interest related to this study.