Submission to Journal of Clinical Periodontology: CPE-12-12-4284.R2
TITLE
Disparity in embrasure fill and papilla height between tooth- and implant-borne fixed restorations in the anterior maxilla: a cross-sectional study
RUNNING TITLE
Embrasure fill and papilla height
KEY WORDS
Dental implant, tooth, fixed partial denture, embrasure, papilla
AUTHORS
Cosyn, J., Raes, M., Packet, M., Cleymaet, R., De Bruyn, H.
AFFILIATIONS AND INSTITUTIONS
Jan Cosyn1,2, Magalie Raes1, Mathieu Packet3, Roberto Cleymaet2,3, Hugo De Bruyn1,4
1University of Ghent, Faculty of Medicine and Health Sciences, Dental School, Department of Periodontology and Oral Implantology, De Pintelaan 185, B-9000 Ghent, Belgium
2Free University of Brussels (VUB), Faculty of Medicine and Pharmacy, Dental Medicine, Laarbeeklaan 103, B-1090 Brussels, Belgium
3Private multidisciplinary practice, Ghent, Belgium
4Malmö University, Faculty of Odontology, Department of Prosthodontics, 205 06 Malmö, Sweden
CONTACT ADDRESS CORRESPONDING AUTHOR
Jan Cosyn
University of Ghent, Faculty of Medicine and Health Sciences, Dental School, Department of Periodontology and Oral Implantology, De Pintelaan 185, B-9000 Ghent, Belgium
E-mail:
WORD COUNT
3003
CONFLICT OF INTERESTS AND SOURCE OF FUNDING
The authors declare they have no conflict of interests. The study was supported by the authors and their institutions.
ABSTRACT
Purpose: The objective of the present study was to compare interproximal fill and papilla height between different embrasures.
Material and methods:One hundred and fifty non-smoking consecutivepatients (mean age 54, range 32–73; 63 males and 87 females) without periodontal disease were selected in a multidisciplinary practice during regular supportive care. All had been treated for multiple tooth loss in the anterior maxilla at least one year earlier by means of a fixed restoration on teeth (n = 50) or implants (n = 100)using straightforward procedures (without hard and/or soft tissue augmentation). Embrasure fill was assessed by means of Jemt’s papilla index and papilla height was registered following local anesthesia by means of bone sounding by one clinician.
Results:Tooth-pontic and tooth-implant embrasures demonstrated comparable interproximal fill and papilla height (≥ 58 % Jemt’sscore 3; mean papilla height ≥ 4.1 mm). Between missing teeth, embrasure fill and papilla height were lowerregardless of the embrasure type. The implant-implant and implant-ponticembrasure demonstrated comparable outcome (≤ 42 % Jemt’s score 3; mean papilla height ≤ 3.3 mm; p ≥ 0.416), which was significantly poorer when compared to the pontic-ponticembrasure (82 % Jemt’sscore 3; mean papilla height 3.7 mm; p ≤ 0.019). Overall, papilla index and papilla height demonstrated a weak correlation (Spearman’s correlation coefficient: 0.198; p = 0.002).
Conclusions:The reestablishment of a papilla is difficult when there is no tooth involved. In that scenario a short papilla should be expected and implant-borne restorations demonstrate the poorest outcome. Moreover, an implant with a ponticmay not perform better than adjacent implants.
CLINICAL RELEVANCE
Scientific rationale:The objective of the present study was to compare interproximal fill and papilla height between different embrasures in 150 consecutivepatients with fixed restorations on teeth or implants.
Principal findings:The reestablishment of a papilla was most difficult between missing teeth. The implant-implant and implant-ponticembrasure demonstrated comparable outcome, which was significantly poorer when compared to the pontic-ponticembrasure.
Practical implications:An implant with a pontic may not perform better than adjacent implants in terms of papilla presence.
Introduction
The aesthetic outcome of a prosthetic treatment is difficult to evaluate as it may be affected by many parameters. In a recent systematic reviewon the professional assessment of aesthetics in implant dentistry this complexity was highlighted and the need for a consensus on objective and well-defined aesthetic parameters was described (Benic et al. 2012). Albeit there is currently little agreement on such parameters, the presence of papillae may be considered an important characteristic of an aesthetic result.
A number of studies have been published on papillae in the context of single implant treatment (Grunder 2000, Choquet et al. 2001, Kan et al. 2003, HenrikssonJemt 2004, Cardaropoli et al. 2006, Juodzbalys & Wang 2007, Lai et al. 2008, Chen et al. 2009, Cosyn et al. 2011, 2012a,Buser et al. 2011, den Hartog et al. 2008, 2011a,b, Gallucci et al. 2011, Raes et al. 2011). Choquet et al. (2001) described complete papilla index in 58 % of the cases between a tooth and an implant with mean papilla height of nearly 4 mm as assessed by bone sounding. The latter may not differ substantially from papilla height between teeth (Tarnow et al. 1992), which has been explained by the preservation of bone and supracrestal fibers at the tooth facing the implant restoration. Hence, factors causing loss of these tissues such as periodontal disease, multiple surgeries with papilla openings and incorrectimplant positioning in relation to the tooth, may in turn be responsible forpartial or complete papilla loss (Cosyn et al. 2012c).
A recent study has shown that a conventional bridge is still a common treatment concept for single tooth replacement in general practice (Cosyn et al. 2012b). In this concept, the embrasure is formed by a tooth and a pontic. The tissue preserving effect of a tooth may also apply here, which is in line with a‘predictable papilla length’ of about 5 mm for interproximal tooth surfaces as described by Salama and co-workers (1998).On the other hand,data on papilla index for a tooth-pontic(T-P) embrasure are lacking. Hence, the first objective was to compare papilla index and height between the tooth-implant (T-I) and the T-Pembrasure.
Complexity increases when two adjacent teeth need replacement in the anterior maxilla. Especially the re-establishment of a papilla between missing teeth is particularly challenging since the interdental bone peak will flatten following tooth loss. Essentially, such an embrasure can be formed by two pontics, two implants or an implant and a pontic.Salama and co-workers (1998) observed a ‘predictable papilla length’ of about 5 mm for interproximal tooth surfaces, 4.5 mm between adjacent implants and 5.5 mm for interproximal implant surfaces not facing another implant. These data suggest the most optimal papilla for an implant-pontic(I-P) and pontic-pontic(P-P) embrasure. However, the lack of clear selection criteria and the fact that papilla length was registered by sounding the first tooth- or implant-to-bone contact instead of the interproximal crestal bone peak, make a correct interpretation difficult.Tarnow and co-workers (2003) described mean papilla height of 3.4 mm between implants as assessed by soundingthe interproximal crestal bone peak on the basis of a quite heterogeneous sample in terms of implant position (anterior as well as posterior), implant type (one- as well as two-piece implants) and bone conditions (native as well as regenerated bone). To overcome aesthetic failures as a result of a deficient papilla, Tarnow et al. (2003) and Barreto et al. (2008) suggested to place only one implant instead of two and to splint it to a cantilevered ovate pontic.Interestingly however, Tymstra et al. (2011) demonstrated identical embrasure fill for the implant-implant (I-I) and I-Pembrasure. Limitations include a very small study sample (5 patients per embrasure type) and lack of data on papilla height. This information clearly shows that the available literature falls short when it comes to data on papillae formed by two pontics, two implants or an implant and a pontic. Hence, another objective was to compare papilla index and height between these embrasures. This information is important from an aesthetic point of view (Tarnow et al. 1992, Jemt 1997)and may become one of the reasons to select or refute a prosthetic treatment option in clinical practice.
Material and Methods
Patient selection
This cross-sectional study was based on a convenience sample of 150 consecutivepatients consulting for regular supportive care in a multidisciplinary practice in Ghent between September 2009 and January 2012. Patients were selected on the basis of specific inclusion and exclusion criteria.
Inclusion criteria were as follows:
-Fixed restoration on teeth (FRT) or implants (FRI) replacing 2 or more adjacent teeth in the anterior maxilla (15-25) and at least 1 year in function.
-Prosthetic procedures performed by 1 of 2 restorative dentists (RC/MP).
-With respect to FRI, surgical procedures performed by 1 of 2 implant surgeons only using two-piece implant systems (JC/MP).
-Informed consent.
Exclusion criteria were as follows:
-Systemic diseases.
-Intake of medication that could induce gingival overgrowth.
-Smoking or history of smoking.
-Prostheses without ovate pontic design.
-History of periodontal treatment.
-Periodontal disease at the time of examination.
-Hard and/or soft tissue augmentation procedures.
-With respect to FRI, suboptimal implant position (T-I distance < 2 mm, I-I distance < 3 mm) as clinically and/or radiographically assessed.
-With respect to FRI, peri-implant probing depth > 4 mm and/or bone loss ≥ 2 mm.
Measurements between an abutment tooth and adjacent pontic (T-P) and between pontics (P-P) were performed in 50 patients with a FRT (Group I). Measurements between a tooth and an implant (T-I) and between adjacent implants (I-I) were performed in 50 patients with a FRI (Group II). In another 50 patients with a FRI, papillae were studied between an implant and a pontic (I-P) (Group III).All embrasures were evaluated in the anterior maxilla and if the embrasure type of interest was present at multiple locations in a patient, the most central location was selected. The study protocol was approved by the Ethical Committee of the University Hospital in Ghent.
Outcome variables
The following outcome variables were registered for 250 embrasures with different embrasures in 150 patients by the same experienced and calibrated clinician (JC):
-Embrasure fill was assessed by means of Jemt’s papilla index(Jemt 1997).
-Papilla height was registered following local anesthesia by means of bone sounding using a periodontal probe (PCPUNC 15, Hu-Friedy®, Leimen, Germany) as described by Tarnow et al. (1992). All recordings were rounded off to the nearest 0.5 mm.
Calibration for papilla index was based on 20 duplicate recordings in 4 patients with a time interval of 1 week.Calibration for papilla height was based on a comparison of clinical height to radiographic height of 10 central maxillary papillae in 10 patients. Radiographic papilla height was determined on the basis of digital radiographs taken with the long-cone paralleling technique. Designated software (DBSWIN, Dürr Dental AG, Bietigheim-Bissingen,Germany) was used to calculate radiographic papilla height. Cases used for calibration included all embrasure types.
Statistical analysis
Data analysis was performed with the patient as the experimental unit. The majority of the data were non-normally distributed and therefore non-parametric tests were applied. For all continuous variables (age, years in function, papilla height) mean values and standard deviations were calculated, whereas frequency distributions were made for categorical variables (gender, papilla index). Disparities between patient groups in terms of age and years in function were evaluated using the Kruskal-Wallis test. For gender, the Fisher’s exact test was adopted. The Wilcoxon signed ranks test was used to study disparities in papilla index and papilla height between paired data (T-P and P-P within the same patient; T-I and I-I within the same patient). The Fisher’s exact test was adopted to compare unpaired data in terms of papilla index (T-P and T-I in different patients; P-P, I-I and I-P in different patients). Disparities in papilla height between unpaired data were evaluated using the Mann-Whitney test (T-P and T-I in different patients) or Kruskal-Wallis test (P-P, I-I and I-P in different patients).The level of significance was set at 0.05 with no correction for multiple testing.
Results
The total sample included 150 patients with a mean age of 54 (range 32-73). Sixty-three were males and 87 were females. Restorations had been on average 4 years in function (range 1-10). Table 1 gives an overview of patient’s age, gender and years in function of the restoration per patient group. Patient groups did not differ in any of these aspects (p ≥ 0.091).Sixty-seven patients were rehabilitated using parallel-walled Biomet 3i®(Palm Beach, Florida, USA) implants with an external connection, 22 using tapered Biomet 3i® implants with an internal connection and 11 using Replace Select tapered TiUnite® (Nobel Biocare, Göteborg, Sweden) implants.
(HERE APPROXIMATELY TABLE 1 PLEASE)
Embrasure fill
Duplicate registration of papilla index resulted in high intra-examiner agreement (Spearman’s correlation coefficient: 0.915; p ≤ 0.001).
Table 2 shows frequency distributions on papilla index sorted per embrasure type. T-P and P-P embrasures assessed in the FRT group, did not differ significantly in papilla index (p = 0.439). In the FRI group with adjacent implants however, significantly less interproximal fill was observed for I-I when compared to T-I embrasures(p < 0.001). Comparison of unpaired data in different patient groups revealed comparable papilla indexfor T-P and T-I embrasures (p = 0.185).Particularly interesting from an aesthetic point of view was the evaluation of papilla index between missing teeth with P-P, I-I or I-P embrasures as possible solutions. Significantly more interproximal fill was observed for the P-P embrasure when compared to I-I (p < 0.001) and I-P (p < 0.001). I-I and I-P embrasures demonstrated comparable papilla index (p = 0.416). As also shown in table 2, complete interproximal fill was only accomplished in less than half of the cases when one or two implants formed the embrasure. A clinical example of an I-P embrasure is shown in figure 1.
(HERE APPROXIMATELY TABLE 2 & FIGURE 1PLEASE)
Papilla height
Clinical and radiographic papilla height demonstrated high agreement (Spearman’s correlation coefficient: 0.994; p ≤ 0.001).Based on the total patient sample papilla height and years in function demonstrated a significant, yet weak correlation (Spearman’s correlation coefficient: 0.178; p = 0.029).
Table 3 shows the results on papilla height sorted per embrasure type. T-P and P-P embrasures as assessed in the FRT group, differed significantly in papilla height with papillae being on average 0.5 mm shorter between pontics (p = 0.001). In the FRI group with adjacent implants, papillae were on average 0.8 mm shorter for I-I when compared to T-I embrasures (p < 0.001). Comparison of unpaired data in different patient groups revealed comparable papilla height for T-P and T-I embrasures (p = 0.383). When scrutinizing possible solutions for recreating papillae between missing teeth, papillae were on average 0.4 – 0.5 mm higher for the P-P embrasure when compared to I-I (p = 0.019) and I-P (p = 0.004). I-I and I-P embrasures demonstratedcomparable papilla height (p = 0.486). As also shown in table 3, the probability of ending up with an ideal papilla (≥ 4 mm) between missing teeth was reduced by half when one or two implants formed the embrasure.
(HERE APPROXIMATELY TABLE 3 PLEASE)
Correlation of papilla indexwith papilla height
On the basis of the total sample (n = 250) papilla index and papilla height demonstrated a weak, yet significant correlation (Spearman’s correlation coefficient: 0.198; p = 0.002).
Discussion
The amount of soft tissues filling the embrasure may be affected by ample factors. These have been identified in a study on single implant treatment using multivariate analyses and include periodontal disease, multiple surgeries with papilla openings and factors relating to the implant and contact point position (Cosyn et al. 2012c). In the present study patient selection was very strict taking into account most of these factors. Albeit this stringent intake took many years, it ensured a quite homogeneous study sample allowing for comparisons of multiple embrasures in different patient groups. Salama and co-workers (1998) were the first to document ‘predictable papilla length’ for different embrasures. However, their findings are difficult to interpret given the lack of clear selection criteria and the fact that papilla length was registered by sounding the first tooth- or implant-to-bone contact instead of the interproximal crestal bone peak. Clearly, accurate data on papillae for different embrasures are needed as these may be important from an aesthetic point of view.
A key finding was that the T-I and T-Pembrasure demonstrated similar papilla index(≥ 58 % with complete fill) and papilla height (≥ 4.1 mm on average). These data do not differ substantially from those between teeth (Tarnow et al. 1992), suggesting a tissue preserving effect of the tooth. On the other hand, when the embrasure was not formed by a tooth significantly less embrasure fill and/or papilla height were found. This situation occurs when adjacent teeth are missing and need replacement. The most favourable outcome was found when such an embrasure was formed by two pontics (82 % with complete fill and 3.7 mm mean papilla height). Whenever one or two implants formed such an embrasure, significantly less embrasure fill (≤ 42 % with complete fill) and papilla height (≤ 3.3 mm on average) were demonstrated.Tarnow et al. (2003) described mean papilla height of 3.4 mm between implants, which is very similar to our observation (3.3 mm on average). Interestingly, the I-Pembrasure did not perform better than the I-Iembrasure in contrast to previous beliefs (Tarnow et al. 2003, Barreto et al. 2008), which is in agreement with a recent pilot study (Tymstra et al. 2011). This information is clinically relevant when it comes to replacing two missing adjacent teeth in the anterior maxilla. The installation of only one implant with a cantilevered pontic is less expensive than two implants, but may not result in superior aesthetics as shown in this study. Moreover, clinicians should be aware of the fact that there are only limited data on implant success and complications for a cantilever bridge on only one implant (Aglietta et al. 2009, 2012,Tymstra et al. 2011).
By and large, about 1 mm shorter papillae were demonstrated between implants or between an implant and a pontic when compared to embrasures formed by one or two teeth. Possibly this is related to a lack of inserting supracrestal fibers around implants(Berglundh et al. 1991). In addition, one may also expect significant loss in the vertical dimension in such embrasures since the bone peak will flatten following tooth extraction. The papillary asymmetry may therefore even become 2 mmas described by Kourkouta et al. (2009), which is detrimental from an aesthetic point of view when a central and lateral incisor or a lateral incisor and a cuspid on one side need replacement. Ridge preservation, orthodontic extrusion, bone augmentation, distraction osteogenesis and root submergence have been proposed as treatment options (Zetu & Wang 2005, Salama et al. 2007). Given the technical complexity of these concepts and their single focus on bone foundation, it is clear that implant dentistry has no straightforward solution for this problem and a FRT with a P-Pembrasure may be considered the best option from an aesthetic point of view to treat this. On the other hand, the range on papilla height for a P-Pembrasure was very high in this study and one could question whether it is ethical to remove sound tooth structure in order to install a FRT. Future studies should therefore focus on papilla reconstruction between implants or between an implant and a pontic using connective tissue grafting. Since it is also unclear whether further improvement in the papilla formed by two pontics is possible using connective tissue grafting, prospective studies should also focus on FRT.