Zurdo Jose Cristina Roma o Jan L.

Wennstro m

Survival and complication rates of implant-supported xed partial dentures with cantilevers: a systematic review

Authors afliations: Zurdo, Cristina Roma Jose o, Institute for Postgraduate Dental Education, University of Central Lancashire, UK Jan L. Wennstro m, Department of Periodontology, Institute of Odontology, The Sahlgrenska Academy at University of Gothenburg, Sweden Correspondence to: Zurdo Jose Institute for Postgraduate Dental Education Greenbank Building, Room 304 University of Central Lancashire Preston PR1 2HE UK Tel.: 44 116 270 87 52 Fax: 44 116 270 0664 e-mail: jose.zurdo@btopenworld.com
Conicts of interest: The authors declare no conicts of interest.

Key words: cantilever extensions, complications, xed partial dental prosthesis, implantsupported, survival, systematic review Abstract Objective: The objective of the present systematic review was to analyze the potential effect of incorporation of cantilever extensions on the survival rate of implant-supported xed partial dental prostheses (FPDPs) and the incidence of technical and biological complications, as reported in longitudinal studies with at least 5 years of follow-up. Methods: A MEDLINE search was conducted up to and including November 2008 for longitudinal studies with a mean follow-up period of at least 5 years. Two reviewers performed screening and data abstraction independently. Prosthesis-based data on survival/ failure rate, technical complications (prosthesis-related problems, implant loss) and biological complications (marginal bone loss) were analyzed. Results: The search provided 103 titles with abstract. Full-text analysis was performed of 12 articles, out of which three were nally included. Two of the studies had a prospective or retrospective casecontrol design, whereas the third was a prospective cohort study. The 5year survival rate of cantilever FPDPs varied between 89.9% and 92.7% (weighted mean 91.9%), with implant fracture as the main cause for failures. The corresponding survival rate for FPDPs without cantilever extensions was 96.396.2% (weighted mean 95.8%). Technical complications related to the supra-constructions in the three included studies were reported to occur at a frequency of 1326% (weighted mean 20.3%) for cantilever FPDPs compared with 012% (9.7%) for non-cantilever FPDPs. The most common complications were minor porcelain fractures and bridge-screw loosening. For cantilever FPDPs, the 5-year event-free survival rate varied between 66.7% and 79.2% (weighted mean 71.7%) and between 83.1% and 96.3% (weighted mean 85.9%) for noncantilever FPDPs. No statistically signicant differences were reported with regard to peri-implant bone-level change between the two prosthetic groups, either at the prosthesis or at the implant level. Conclusion: Data on implant-supported FPDPs with cantilever extensions are limited and therefore survival and complication rates should be interpreted with caution. The incorporation of cantilevers into implant-borne prostheses may be associated with a higher incidence of minor technical complications.

Date: Accepted 20 May 2009

To cite this article: Zurdo J, Roma o C, Wennstro m JL. Survival and complication rates of implant-supported xed partial dentures with cantilevers: a systematic review. Clin. Oral Impl. Res. 20 (Suppl. 4), 2009; 5966. doi: 10.1111/j.1600-0501.2009.01773.x

The selection of prosthetic options to replace missing teeth should be based on scientic evidence. The incorporation of cantilever extensions into implant-borne reconstructions may be considered as an

option in situations where local conditions of the residual edentulous ridge preclude the possibility to place an implant. However, it has been claimed that cantilever extensions increase the risk of bending

 c 2009 John Wiley & Sons A/S

59

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

overload and that this in turn may compromise the prognosis of the prosthetic rehabilitation (Rangert et al. 1995). In a series of recent systematic reviews, the information available in the literature on the success/survival rates and the incidence of biological and technical complications of different designs of tooth and implant-supported xed prosthesis was summarized (Lang et al. 2004; Pjetursson et al. 2004a, 2004b, 2007; Tan et al. 2004; Jung et al. 2008; Pjetursson & Lang 2008). These reviews revealed that the incidence of technical complications was signicantly higher for implant-supported than for tooth-supported prostheses. For toothsupported prostheses, technical complications were found to be more frequent for cantilever than for end-abutment prostheses. However, the extent to which cantilevers may affect the survival and complication rates of implant-supported xed dental prostheses (FDPs) was not analyzed in these reviews. Long-term clinical studies have demonstrated that implant-supported full-arch reconstructions with bilateral cantilevers in the mandible exhibited high survival rates (Adell et al. 1981, 1990; Albrektsson et al. 1988). However, it has been suggested that certain cantilever lengths may decrease the survival of the prostheses (Shackleton et al. 1994). There are inherent biomechanical differences in the implant treatment of completely edentulous arches and posterior partially edentulous segments, as the partial prosthesis does not benet from crossarch stabilization and, therefore, is more susceptible to bending loads (Rangert et al. 1997). In vitro studies revealed that implantsupported cantilever prostheses lead to a high stress concentration at the marginal bone level of the implants, particularly at the implant closest to the cantilever extension (Sertgoz & Guvener 1996; Stegaroiu et al. 1998; Zampelis et al. 2007), which was considered to pose a risk for marginal bone loss at the implants. On the other hand, experimental studies in the dog model showed that excessive static loading of implants did not result in marginal bone loss or loss of osseointegration, but that the bone tissue adjacent to the loaded implants exhibited a greater density compared with unloaded implants (Gotfredsen et al. 2001a, 2001b). Implant loss as a result of

excessive occlusal load in a lateral direction was demonstrated in experimental studies in a monkey model (Isidor 1996, 1997), suggesting that it is possible to induce loss of osseointegration to the implant, but not marginal bone loss, when forces are beyond the repair potential of the bone. However, in humans, the biological impact of excessive load remains unclear. The focus question for the current systematic review was To what extent do cantilevers affect survival and complications of implant-borne reconstructions in the partially dentate patient. Hence, the aims dened were to analyze the potential effect of the incorporation of cantilever extensions on (i) the survival (or failure) rate of implant-supported xed partial dental prostheses (FPDPs) and (ii) the incidence of technical and biological complications, as reported in longitudinal studies with at least 5 years of follow-up.

studies should include prosthesis-based data on success, survival or loss rate, technical complications (prosthesis-related problems, implant loss) and/or biological complications (marginal bone loss).
Exclusion criteria

Letters, experimental studies and narrative reviews were explicitly excluded. Studies with o10 patients examined at the end of follow-up, and studies focusing on overdentures were also excluded. Studies from which data on selected outcome variables could not be directly retrieved or calculated were not considered.
Selection of studies

Materials and methods

Search strategy

A protocol to be followed was agreed upon by the authors before the initiation of the literature search. Anticipating very few, if any, randomized clinical trials (RCT) related to the focused topic, the decision was taken to use a broad search strategy. An initial electronic search on MEDLINE (PubMed) from 1966 up to and including November 2008 was conducted for English-language articles published in the dental literature, using the keywords dental implants AND cantilever(s). The search yielded 103 references to be screened for possible inclusion based on titles and abstracts.

Inclusion criteria

Systematic reviews and longitudinal prospective/retrospective studies (RCT, controlled clinical trials and cohort studies) reporting data with regard to the outcome of treatment with implant-supported FPDPs with cantilever extensions after a mean function time of at least 5 years were accepted for inclusion. The selection of a function time of at least 5 years was based on the consensus from a previous workshop regarding the recommended follow-up time for evaluation of implant therapy (Wennstro m & Palmer 1999). Further, the

Two independent reviewers (JZ and CR) screened the 103 abstracts retrieved from the electronic search for possible inclusion in the review. Ten abstracts were accepted for inclusion by both reviewers, and further three by just one reviewer. After discussion, a consensus was reached to include one of the latter publications and reject the other two articles. The Kappa score for agreement between the reviewers for screening of abstracts was 0.85 (Fig. 1). Full-text articles were obtained of the 11 selected publications. In addition, hand searches were performed on bibliographies of the selected articles as well as of identied narrative reviews. Four publications were found that reported data on subject samples that were included in the already identied studies. Such repeated reports were grouped together. One further article was identied for inclusion after the hand search. The two reviewers independently assessed the 12 full-text articles to determine whether they fullled the dened criteria for nal inclusion. Any disagreement was resolved by discussion. Three studies were found to qualify for inclusion in the review, while nine studies had to be excluded (Fig. 1).
Excluded studies

Out of the nine studies that were excluded following full-text analysis, seven had a mean follow-up less than 5 years, including two studies focusing on complete FDPs, and in further two studies selected outcomes could not be retrieved (Table 1).
Data extraction

Data were extracted independently by the two reviewers using a data extraction form
 c 2009 John Wiley & Sons A/S

60 |

Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

Initial electronic search 103 titles/abstracts

Independent screening by 2 reviewers

(mm or crown units); implant system used, number and length of implants involved in the FPDPs; and type of antagonist dentition. Technical complications were divided into three categories:

Kappa score 0.85 10 abstracts agreed by both reviewers 3 to discuss: 1 abstract included, 2 rejected 11 abstracts selected for full-text review Further hand-searching; 1 article included for full-text review (*) 12 selected for full-text review

i. Lost reconstructions i.e. reconstructions not in situ at the follow-up examination. ii. Lost implants (including reasons for the loss). iii. Complications related to the suprastructure fractures or deformations of the framework or veneers, loss of retention and screw or abutment loosening.  Biological complications marginal bone loss at the FDP level and at the implant next to the cantilever, respectively.

Excluded 9 articles

3 studies included in review

Studies in which data on a certain variable were lacking or could not be calculated were scored as not reported for the variable in question. Weighted mean values and the 95% condence interval (CI) for the various outcomes were calculated.

(*) Five further publications related to 5 of the originally selected were also included to retrieve further data, but were not counted towards the final number of included papers as they reported on the same patient material.
Fig. 1. Process of identifying the studies included in the review, from an initial 103 titles/abstracts.

Results
The literature search conrmed the inexistence of publications on RCT comparing the outcomes of implant-supported FPDPs with and without cantilever extensions.
Characteristics of the included studies (Table 2)

Table 1. Excluded articles (cohort studies) after full-text examination and reasons for exclusion
Reference Blanes et al. (2007) Reason for exclusion Implant-based data analysis Information regarding number of prosthesis with cantilever extensions, prostheses survival and/or complications not available Mean follow-up o5 years Mean follow-up o5 years, selected outcomes not retrievable Mean follow-up o5 years Mean follow-up o5 years Mean follow-up o5 years Mean follow-up o5 years, cantilevered bridges were all xed complete dental prosthesis (FCDP) Cross-sectional study design Mean follow-up o5 years, focus on FCDP

Nedir et al. (2006) Tawil et al. (2006) Becker (2004) Romeo et al. (2003) Johansson & Ekfeldt (2003) Kucey (1997)

Ranger et al. (1995) Shackleton et al. (1994)

previously agreed upon. Disagreement regarding data extraction was resolved by discussion and consensus. The following variables were recorded:
 c 2009 John Wiley & Sons A/S

Number of subjects included at baseline and at the follow-up examination; number and characteristics of the FPDPs; extension of cantilever segment

The study characteristics of the three publications that qualied for inclusion (Wennstro m et al. 2004; Kreissl et al. 2007; Ha lg et al. 2008) are presented in Table 2. Two of the studies had a prospective or a retrospective casecontrol design, whereas the third was a prospective cohort study. In one study (Wennstro m et al. 2004) all included patients had a 5-year follow-up, whereas in the other two studies the followup time among the patients ranged between 0 and 12.7 years, with a reported mean follow-up time of 55.3 years. The implant systems used in the studies were ITI (Straumann AG, Waldenburg, Switzerland, one study), Astra (Astra Tech AB, Mo lndal, Sweden, one study) and 3i (Implant Innovation Inc., West Palm Beach, FL, USA, one study).
Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

61 |

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

Table 2. Characteristics of the included studies

Reference Type of study Number of Implant Follow-up Number of patients system period reconstructions in years (range) Cantilever Mean extension number of implants in prosthesis Mean length of implants (range) Cant: 1.7 Non-C: 1.2 10.1 mm (612) Type of Comments antagonists

Casecontrol studies lg et al. Ha Retrospective 54 (2008)

ITI

Mean 5.3 years (312.7)

Cant: 27 Non-C: 27

Teeth or 1 crown FPDPs unit (26 prostheses) on teeth 2 crown units (1 prosthesis)

Cant: 8 of 27 patients with single implant Non-C: 22 of 27 patients with single implant

m Prospective Wennstro et al. (2004)

45

Astra Tech

5 years

Cant: 24 Non-C: 26

Cant: 2.6 Non-C: 2.8 12.7 mm (819 mm)

Mean 9 mm

Teeth or FPDPs on teeth except 1 (implantsupported FPDP) NR Non-C: 46 of 89 patients with single implant

Cohort studies Kreissl et al. Prospective (2007)

76

3i

Mean 5 years (080 months)

Cant: 23 Non-C: 89

Cant: 2.7 Non-C: 1.6 NR

NR

NR, not reported; Cant, cantilever FPDP; Non-C, non-cantilever FPDP; FPDP, xed partial dental prostheses.

The total number of prostheses included in the three studies was 216 (74 with and 142 without cantilever extensions), with the number of cantilever prostheses varying between 23 and 27 among the studies. The mean number of implants supporting the prostheses ranged between 1.7 and 2.7 for cantilever FPDPs and between 1.2 and 2.8 for non-cantilever FPDPs. In the study by Ha lg et al. (2008), eight (30%) of the 27 included cases in the cantilever FPDP group had only one implant supporting the cantilever reconstruction. Two of the studies (Kreissl et al. 2007; Ha lg et al. 2008) included a large proportion of single-implant restorations in the non-cantilever group, 52% and 81%, respectively. Data on the mean length of implants and cantilever extensions were available in two studies (Wennstro m et al. 2004; Ha lg et al. 2008). The mean overall length of the implants varied between 10.1 and 12.7 mm, while the length of the cantilever extension was described as crown units in one study (one unit in all but one FPDP) or as a mean length in millimeters in the other (9 mm). The two casecontrol studies also provided information regarding the type of antagonist dentition, which,

in all cases but one (implant supported FPDP), was natural teeth or tooth-supported FPDPs.
Loss of prostheses and implants (Table 3)

Information regarding failure (loss) rates for prostheses with and without cantilevers could be retrieved from all three included studies. The loss of cantilever prostheses ranged between 4.3% and 11.1% (weighted mean 8.1%, 95% CI 4.212) compared with 3.74.5% (4.2%, 95% CI 3.74.7) for FPDPs without cantilevers. Hence, the calculated weighted mean 5-year survival rate was 91.9% (95% CI 8895.8) for cantilever FPDPs and 95.8% (95% CI 95.396.3) for FPDPs without cantilevers. The reasons for loss of the prostheses in the cantilever groups in the two casecontrol studies (Wennstro lg et al. m et al. 2004; Ha 2008) were reported to be fracture of an implant (three cases) and need for remaking the supra-construction (one case), while the two events in the non-cantilever groups were due to implant fracture. In the cohort study (Kreissl et al. 2007) biological complications were indicated as the reason for the loss of prostheses (one cantilever and four non-cantilever prostheses).

Implant loss during the 5-year function period was reported in all three studies. In each of the two casecontrol studies the implant loss was two (3.2%/4.3%) for the cantilever group and one (1.4%/3.1%) for the non-cantilever group, all due to fractures. The corresponding gures for the cantilever and non-cantilever FDP groups in the cohort study (Kreissl et al. 2007) were one (1.6%) and four (2.8%), respectively, and all were reported to have been lost as a consequence of biological complications. The weighted mean 5-year rate of implant loss was 2.9% (95% CI 1.44.5) for cantilever FPDPs compared with 2.4% (95% CI 1.43.5) for non-cantilever FPDPs.
Technical complications supra-constructions

Technical complications related to the supra-constructions in the three included studies were reported to occur at a frequency of 1326% (weighted mean 21.6%, 95% CI 11.931.2) for cantilever-FPDPs compared with 012% (10.3%, 95% CI 2.118.5) for non-cantilever FPDPs. The most common complications were minor porcelain fractures and bridge-screw loosening. Wennstro m et al. (2004) reported a total of six incidences (three bridge-screw loosening
 c 2009 John Wiley & Sons A/S

62 |

Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

Table 3. Technical and biological complications implant-supported FPDPs

Reference Technical complications 5 years Lost prostheses n (%) Biological complications 5 years Comments Lost implants Complications Marginal bone loss (mean mm) n (%) supraconstruction Prosthesis Implant n (%) next to cantilever Cant: 2 (4.3%) Non-C: 1 (3.1%) Cant: 6 (22%) Non-C: 0 (0%) Cant: 0.23 Non-C: 0.09 Cant: 0.23 Non-C: 0.05 The losses of prostheses were due to implant fractures (2 Cant and 1 NonC) and need for remaking the supraconstruction (1 Cant). All lost implants were due to fracture No signicant difference in bone loss Cant vs. Non-C 33% smokers in Cant, 11% in Non-C Baseline: prosthesis installation Peri-implantitis affected 4 implants (1 Cant vs. 3 Non-C) The losses of prostheses were due to implant fractures All lost implants were due to fracture No signicant difference in bone loss Cant vs Non-C 42% smokers in Cant, 19% in Non-C Baseline: prosthesis installation Peri-implantitis NR All lost prostheses were due to implant loss Reasons for loss of implants not reported % smokers not reported No radiographic analysis Peri-implantitis NR Event free survival rate: Cant: 71.7% (64.279.1) Non-C: 85.9% (77.794.1)

Casecontrol studies lg et al. Cant: 3 Ha (2008) (11.1%) Non-C: 1 (3.7%)

m Cant: 2 Wennstro et al. (8.3%) (2004) Non-C: 1 (3.8%)

Cant: 2 (3.2%) Non-C: 1 (1.4%)

Cant: 3 (13%) Non-C: 3 (12%)

Cant: 0.49 Non-C: 0.38

Cant : 0.39 Non-C: 0.23

Cohort studies Kreissl Cant: 1 et al. (4.3%) (2007) Non-C: 4 (4.5%)

Cant: 1 (1.6%) Non-C: 4 (2.8%)

Cant: 6 (26%) Non-C: 11 (12%)

NR

NR

Weighted mean (95% CI)

Cant: 8.1% (4.212) Non-C: 4.2% (3.74.7)

Cant: 2.9% (1.44.5) Non-C: 2.4% (1.43.5)

Cant: 20.3% (12.827.9) Non-C: 9.7% (1.917.6)

Cant: 0.35 (0.10.6) Non-C: 0.23 ( 0.10.6)

Cant: 0.31 (0.10.5) Non-C: 0.14 (00.3)

NR, not reported; Cant, cantilever FPDP; Non-C, non-cantilever FPDP; FPDP, xed partial dental prostheses.

and three minor porcelain fractures), equally distributed between the cantilever and the non-cantilever FDP groups. Ha lg et al. (2008) also reported six events, but all occurring in the cantilever-FDP group (one suprastructure fracture, four minor porcelain fracture, and one re-cementation). Of a total of 17 complications described in the cohort study by Kreissl et al. (2007), six belonged to the cantilever group, corresponding to an incidence rate of 26% as compared with 12% in the non-cantilever group.
Event-free survival rate

64.279.1) and for non-cantilever it was FPDPs 83.1%96.3% (weighted mean 85.9%, 95% CI 77.794.1).

Biological complications (Table 3)

By taking all technical complications reported under consideration, the event-free survival rate over 5 years was calculated. For cantilever FPDPs the event-free survival rate was 66.7%79.2% in the various studies (weighted mean 71.7%, 95% CI
 c 2009 John Wiley & Sons A/S

Bone-level changes were assessed only in the two casecontrol studies (Wennstro m et al. 2004; Ha lg et al. 2008). In both studies, the baseline radiograph was obtained at prosthesis installation and assessments of bone changes were performed at the FDP, implant and site levels. As for the standardization of the radiographic technique, one study (Wennstro m et al. 2004) reported the use of a custom-made stent to optimize the reproducibility of projection geometry, while the other study (Ha lg et al. 2008) used only a standardized parallel long-cone technique.

The mean bone loss at the FDP level in the cantilever and non-cantilever groups in the two studies amounted to 0.23 0.49 mm (weighted mean 0.35 mm, 95% CI 0.100.61) and 0.090.38 mm (weighted mean 0.23 mm, 95% CI 0.05 to 0.52), respectively. The magnitude of bone loss at the implant closest to the cantilever extension (implant level) was 0.230.39 mm (weighted mean 0.31 mm, 95% CI 0.15 0.46) compared with 0.050.23 mm (weighted mean 0.14 mm, 95% CI 0.04 to 0.32) at the end implant or a randomized selected implant in the control group. Potential inuence on peri-implant marginal bone loss of various confounding factors was analyzed in the two studies using multivariate regression analysis. Jaw of treatment (maxilla), and in one of the studies smoking habits (Wennstro m
Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

63 |

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

et al. 2004), was associated with a signicantly increased magnitude of marginal bone loss. Type of FPDPs (with or without cantilever extension) had no signicant effect on marginal bone loss.

Discussion
In the present systematic review, the incidence of loss and technical and biological complications of FDPs with cantilever extensions were described. To address the focus question to what extent do cantilevers affect survival and complications of implant borne reconstructions in partially edentulous patients, the approach in the search was to identify longitudinal studies, prospective or retrospective, with at least 5 years of follow-up and reporting prosthesis-based data. It is generally acknowledged that longitudinal studies with a time span of at least 5 years are required to properly evaluate the outcome of implant treatment (Wennstro m & Palmer 1999; Berglundh et al. 2002; Pjetursson et al. 2004a). Although we accepted for inclusion studies that presented data for a cohort with a mean follow-up of 5 years, only three studies qualied for inclusion. In a previous systematic review on implant therapy (Berglundh et al. 2002), it was suggested to exclude studies in which o80% of the initial subject sample had been followed for 5 years. If this criterion had been applied, only one study (Wennstro m et al. 2004) would qualify for inclusion since in the other two studies (Kreissl et al. 2007; Ha lg et al. 2008) only 4665% of the implant constructions were observed for a period of 5 years. However, because of the scarcity of identied studies addressing the issue of cantilever extensions, we decided to retain these studies. Based on the three studies included in the current review, the calculated overall 5year prosthesis survival rate was 92% for cantilever FPDPs, compared with 96% for non-cantilever FPDPs. Although these gures may indicate a somewhat inferior performance for the cantilever prostheses, the results should be interpreted with caution because the sample size is small. The overall prosthesis survival rate in the included studies was comparable with the results of a previous systematic review on implant-supported FDPs (Pjetursson et al.

2004a), in which a meta-analysis including 14 studies yielded an overall estimated survival rate of 95% (95% CI 92.296.8) after 5 years. The event-free survival rate, i.e. the proportion of reconstructions that remained in function after 5 years without any technical complications, was estimated to be 72% for FPDPs with cantilever extension and 86% for FPDPs without cantilevers. However, the comparison with the available literature is difcult as the way in which data are generated and presented lacks uniformity. In a previous systematic review on implant-supported FDPs (Pjetursson et al. 2004a), including three studies reporting on the incidence of patients without any complication, the estimated success rate after 5 years was 61.3% (95% CI 55.366.8). In another systematic review (Berglundh et al. 2002), the 5-year incidence of technical complications for implant-supported FDPs, calculated at the patient level, was reported to be 25%. Taken together, these data indicate that technical complications are common for implant-supported prostheses, both with and without cantilever extensions. It has been suggested (Pjetursson et al. 2004a) that technical complications should be divided into major (implant fracture, loss of suprastructures), medium (abutment, veneer or framework fracture) and minor (abutment or screw loosening, loss of retention, loss of veneer hole sealing, veneer chipping fracture), and that the type and number of events should be reported separately. Implant fractures were in the two casecontrol studies reported to vary between 1.4% and 4.3% across the treatment groups. Comparable data reported in the systematic review by Pjetursson et al. (2004a, 2004b) on implant-supported FPDPs in general revealed a cumulative incidence of 0.4%. The fact that the four implants that fractured in the cantilever groups of the two casecontrol studies included in the current systematic review had a narrow diameter (3.33.5 mm) suggests that other factors in addition to the cantilever extension should be considered when evaluating the mechanical risks of a specic prosthetic design. The study by Ha lg et al. (2007) had a skewed distribution in the proportion of narrow implants between the treatment groups: 39% and 7% for the cantilever and non-cantilever

groups: respectively. In the other case control study (Wennstro m et al. 2004), one of the two implant fractures in the cantilever group occurred in a patient who was diagnosed as a bruxer. Furthermore, in this study the mean height of the supraconstructions was signicantly greater in the cantilever group than in the non-cantilever group, which could be regarded as yet another potential prosthetic-related factor that might contribute to the incidence of technical complications. Nevertheless, this study reported a low frequency of technical complications (0.12 incidence/ patient) that were equally distributed among the groups and comparable with the calculated 5-year incidence/patient of 0.24 technical complications for FPDs reported in a previous systematic review (Berglundh et al. 2002). Another prosthesis-related factor that has been suggested to contribute to an increase in the rate of technical complications is the type of antagonists (Davis et al. 2003). Further, it is well recognized that loading may be signicantly higher in posterior segments as compared with anterior regions (Rangert et al. 1995) and that more prosthetic complications may occur in the posterior regions (Nedir et al. 2006). In the current review, two studies reported information about implant position and type of antagonists. Eighty-ve percent to 100% of the FPDPs were placed in the premolarmolar regions and all were functioning against natural teeth or FPDPs supported by teeth or implants. The mean overall marginal bone-level change after 5 years at the implant-supported FPDPs reported by two studies included in this review was small and well below the degree of bone loss acceptable according to the success criteria described by Albrektsson et al. (1986). Cantilever extensions did not signicantly inuence the bone changes. On the other hand, factors such as jaw of treatment (maxilla) and smoking habits appeared to be signicant for observed peri-implant marginal bone loss. The presence of cantilever extensions and its effect on crestal bone loss were also analyzed in a prospective study with partially dentate patients who were restored with FPDPs in the posterior region (Blanes et al. 2007). Neither the presence of a mesial nor a distal cantilever had a signicant effect on peri-implant bone
 c 2009 John Wiley & Sons A/S

64 |

Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

loss after an observation period of 6 years. The biological plausibility that excessive loading results in increased marginal bone loss is controversial. Implant failure as a result of excessive occlusal load in a lateral direction was demonstrated in experimental studies in a monkey model (Isidor 1996, 1997), suggesting that it is possible to induce loss of osseointegration when the forces are beyond the repair potential of the bone. However, under clinical conditions, the biological impact and categorization of excessive load remains unclear and its translation into peri-implant marginal bone loss has not been demonstrated. Although early observations made in clinical studies suggested an association between excessive loading and peri-implant bone loss (Quirynen et al. 1992), bivariate analyses should be interpreted with caution because of potential skewed distribution of confounding factors. Other factors such as smoking (Lindquist et al. 1997) and maxilla vs. mandible (Jemt & Lekholm, 1993) have also been associated with peri-implant bone loss. Both casecontrol studies included in this systematic review used multivariate models for analysis and found cantilever extension to be a factor without a signicant effect on observed marginal bone loss when the confounding factors were controlled in the statistical model. Several other variables related to the FPDP conguration have been suggested to inuence its loading capacity, such as the buccolingual occlusal extension of the prostheses in addition to the above-mentioned height of the supra-construction and the number, position and inclination of the supporting implants (Rangert et al. 1997). Also, implant designs and surface roughness inuence the boneimplant support capacity. Hence, determining the potential effect of one loading-associated factor on peri-implant crestal bone loss is complicated unless other variables are adequately controlled in clinical trials.

Peri-implantitis is an important biological complication resulting in bone loss but was not considered as an outcome in the current review because of the lack of a plausible biological relationship with the presence/absence of cantilever extension. Information regarding the condition of the peri-implant tissues in clinical studies related to dental implants is scarce (Berglundh et al. 2002). The incidence of periimplantitis was reported in one study included in the current review (Ha lg et al. 2008); 5.1% of the prostheses were affected. The cumulative incidence of periimplantitis and soft tissue complications for FPDPs after 5 years was estimated to be 8.6% in a previous systematic review (Pjetursson et al. 2004a, 2004b). A more recent systematic review (Zitzmann & Berglundh 2008) described alarmingly higher gures. They retrieved data from two study samples and peri-implantitis was found in these studies in 28% and ! 56% of the subjects. In the current review, only three studies were included and only one study presented data with 480% of the patients followed for 5 years. Of the excluded studies not fullling the criterion of a minimum mean follow-up of 5 years, high short-term survival rates (95.6100%) were reported in three longitudinal studies with a mean follow-up time ranging between 3.9 and 4.5 years (Johansson & Ekfeldt 2003; Romeo et al. 2003; Becker 2004). In another excluded study (Nedir et al. 2006), comparative data of the prosthetic complications associated with different implant prosthetic designs were presented. After a mean follow-up of 3.3 years (44% of the patients followed for 5 years), the complication rate for FPDPs (the majority placed in the posterior regions) was 29.4% for the cantilever and 7.9% for non-cantilever prostheses. It is obvious from this systematic review that there is a need for additional prospec-

tive controlled clinical studies with at least 5 years of follow-up to properly determine to what extent cantilever extensions may affect the survival and complications of implantborne reconstructions.

Conclusions
Within the limitations of the current review, the following conclusions can be made:

The 5-year survival rate was high for both cantilever and non-cantilever FPDPs (91.9% vs. 95.8%). The most common reason for loss of FPDPs with cantilever extensions was implant fracture. The incorporation of cantilevers into implant-borne prostheses was associated with a higher incidence of technical complications related to the supra-constructions (20.3% vs. 9.7% for non-cantilever FPDPs). Minor porcelain fractures and bridge-screw loosening were the most common technical complications. For cantilever FPDPs the 5-year eventfree survival rate was 71.7% compared with 85.9% for non-cantilever FPDPs. The incorporation of cantilevers into implant-borne prostheses did not have any signicant effect on the amount of peri-implant marginal bone loss, either at the prosthesis level or at the implant next to the cantilever.

Clinical implications
An implant-supported FPDP with a short cantilever extension (one tooth unit) is an acceptable restorative therapy, and might be considered as an alternative to procedures that require more advanced surgery (e.g. sinus graft, etc.) or for esthetic reasons.

References meeting the inclusion criteria of the systematic review

Ha lg, G.A., Schmid, J. & Ha mmerle, C.H. (2008) Bone level changes at implants supporting crowns or xed partial dentures with or without cantilevers. Clinical Oral Implants Research 19: 983990. Kreissl, M.E., Gerds, T., Muche, R., Heydecke, G. & Strub, J.R. (2007) Technical complications of implant-supported xed partial dentures in partially edentulous cases after an average observation period of 5 years. Clinical Oral Implants Research 18: 720726. Wennstro m, J., Zurdo, J., Karlsson, S., Ekestubbe, A., Gro ndahl, K. & Lindhe, J. (2004) Bone level change at implant-supported xed partial dentures with and without cantilever extension after 5 years in function. Journal of Clinical Periodontology 31: 10771083.

 c 2009 John Wiley & Sons A/S

65 |

Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

Zurdo et al Survival and complication rates of implant-supported xed partial dentures with cantilevers

References not included in the systematic review, but quoted in the manuscript
nemark, Adell, R., Eriksson, B., Lekholm, U., Bra P.I. & Jemt, T. (1990) A long-term follow-up study of osseointegrated implants in the treatment of totally edentulous jaws. International Journal of Oral & Maxillofacial Implants 5: 347359. nemark, Adell, R., Lekholm, U., Rockier, B. & Bra P.-I. (1981) A 15-year study of osseointegrated implants in the treatment of the edentulous jaw. International Journal of Oral Surgery 10: 387 416. Albrektsson, T. (1988) A multicenter report on osseointegrated oral implants. Journal of Prosthetic & Dentistry 60: 7584. Albrektsson, T., Zarb, G., Worthington, P. & Eriksson, A.R. (1986) The long-term efcacy of currently used dental implants: a review and proposed criteria of success. International Journal of Oral & Maxillofacial Implants 1: 1125. Becker, C.M. (2004) Cantilever xed prostheses utilizing dental implants: a 10-year retrospective analysis. Quintessence International 35: 437 441. Berglundh, T., Persson, L. & Klinge, B. (2002) A systematic review of the incidence of biological and technical complications in implant dentistry reported in prospective longitudinal studies of at least 5 years. Journal of Clinical Periodontology 29 (Suppl. 3): 197212; discussion 232193. Blanes, R.J., Bernard, J.P., Blanes, Z.M. & Belser, U.C. (2007) A 10-year prospective study of ITI dental implants placed in the posterior region. II: inuence of the crown-to-implant ratio and different prosthetic treatment modalities on crestal bone loss. Clinical Oral Implants Research 18: 707714. Davis, D.M., Packer, M.E. & Watson R, .M. (2003) Maintenance requirements of implant-supported xed prostheses opposed by implant-supported xed prostheses, natural teeth, or complete dentures: a 5-year retrospective study. International Journal of Prosthodontics 16: 521523. Gotfredsen, K., Berglundh, T. & Lindhe, J. (2001a) Bone reactions adjacent to titanium implants subjected to static load of different duration. A study in the dog (III). Clinical Oral Implants Research 12: 552558. Gotfredsen, K., Berglundh, T. & Lindhe, J. (2001b) Bone reactions adjacent to titanium implants subjected to static load. A study in the dog (I). Clinical Oral Implants Research 12: 18. Isidor, F. (1996) Loss of osseointegration caused by occlusal load of oral implants. A clinical and radiographic study in monkeys. Clinical Oral Implants Research 7: 143152. Isidor, F. (1997) Histological evaluation of peri-implant bone at implants subjected to occlusal overload or plaque accumulation. Clinical Oral Implants Research 8: 19. Jemt, T. & Lekholm, U. (1993) Oral implant treatment in posterior partially edentulous jaws: a 5year follow-up report. International Journal of Oral & Maxillofacial Implants 8: 635640. Johansson, L.A. & Ekfeldt, A. (2003) Implantsupported xed partial prosthesis. A retrospective study. International Journal of Prosthodontics 16: 172176. Jung, R.E., Pjetursson, B.E., Glauser, R., Zembic, A., Zwahlen, M. & Lang, N.P. (2008) A systematic review of the 5-year survival and complication rates of implant-supported single crowns. Clinical Oral Implants Research 19: 119130. Kucey, B.T. (1997) Implant placement in prosthodontics practice: A ve-year retrospective study. J Prosthet Dent 77: 171176. Lang, N.P., Pjetursson, B.E., Tan, K., Bra gger, U., Egger, M. & Zwahlen, M. (2004) A systematic review of the survival and complication rates of xed partial dentures (FDPs) after an observation period of at least 5 years II. Combined toothimplant supported FDPs. Clinical Oral Implants Research 15: 643653. Lindquist, L.W., Carlsson, G.E. & Jemt, T. (1997) Association between marginal bone loss around osseointegrated mandibular implants and smoking habits: a 10-year follow-up study. Journal of Dentistry Research 76: 16671674. Nedir, R., Bischof, M., Szmukler-Moncler, S., Belser, U.C. & Samson, J. (2006) Prosthetic complications with dental implants: from an up-to-8year experience in private practice. International Journal of Oral & Maxillofacial Implants 21: 919928. Pjetursson, B.E., Bra gger, U., Lang, N.P. & Zwahlen, M. (2007) Comparison of survival and complication rates of tooth supported xed partial dentures and implant supported xed partial dentures and single crowns. Clinical Oral Implants Research 8 (Suppl. 3): 97113. Pjetursson, B.E. & Lang, N.P. (2008) Prosthetic treatment planning on the basis of scientic evidence. Journal of Oral Rehabilitation 35 (Suppl. 1): 7279. Pjetursson, B.E., Tan, K., Lang, N.P., Bra gger, U., Egger, M. & Zwahlen, M. (2004a) A systematic review of the survival and complication rates of xed partial dentures (FDPs) after an observation period of at least 5 years I. implant supported FDPs. Clinical Oral Implants Research 15: 625 642. Pjetursson, B.E., Tan, K., Lang, N.P., Bra gger, U., Egger, M. & Zwahlen, M. (2004b) A systematic review of the survival and complication rates of xed partial dentures (FDPs) after an observation period of at least 5 years IV. Cantilever or extensions FDPs. Clinical Oral Implants Research 15: 667676. Quirynen, M., Naert, I. & van Steenberghe, D. (1992) Fixture design and overload inuence marginal bone loss and xture success in the Branemark system. Clinical Oral Implants Research 3: 104111. Rangert, B., Krogh, P.H., Langer, B. & Van Roekel, N. (1995) Bending overload and implant fracture: a retrospective clinical analysis. International Journal of Oral & Maxillofacial Implants 10: 326334. Rangert, B., Sullivan, R. & Jemt, T. (1997) Load factor control for implants in the posterior partially edentulous segment. International Journal of Oral & Maxillofacial Implants 12: 360370. Romeo, E., Lops, D., Margutti, E., Ghisol, M., Chiapasco, M. & Vogel, G. (2003) Implant-supported xed cantilever prostheses in partially edentulous arches. A seven-year prospective study. Clinical Oral Implant Research 14: 303311. Sertgoz, A. & Guvener, S. (1996) Finite element analysis of the effect of cantilever and implant length on stress distribution in an implant-supported xed prosthesis. Journal of Prosthetic Dentistry 76: 165169. Shackleton, J.L., Carr, L., Slabbert, J.C. & Becker, P.J. (1994) Survival of xed implant-supported prostheses related to cantilever lengths. Journal of Prosthetic Dentistry 71: 2326. Stegaroiu, R., Sato, T., Kusakari, H. & Miyakawa, O. (1998) Inuence of restoration type on stress distribution in bone around implants: a threedimensional nite element analysis. International Journal of Oral & Maxillofacial Implants 13: 8290. Tan, K., Pjetursson, B.E., Lang, N.P. & Chan, E.S.Y. (2004) A systematic review of the survival and complication rates of xed partial dentures (FDPs) after an observation period of at least 5 years. III. conventional FDPs. Clinical Oral Implants Research 15: 654666. Tawil, G., Aboujaoude, N., Younan, R. (2006) Inuence of prosthetic parameters on the survival and complication rates of short implants. Int J Oral Maxillofac Implants 21: 275282. Wennstro m, J. & Palmer, R.M. (1999) Concensus report: clinical trials. In: Lang, N.P., Karring, T. & Lindhe, J., eds. Proceedings of the 3rd European Workshop on Periodontology, 255259. Berlin: Quintessence Publication. Zampelis, A., Rangert, B. & Heijl, L. (2007) Tilting of splinted implants for improved prosthodontic support: a two-dimensional nite element analysis. Journal of Prosthetic Dentistry 97: S35S43. Zitzmann, N.U. & Berglundh, T. (2008) Denition and prevalence of peri-implant diseases. Journal of Clinical Periodontology 35 (Suppl. 8): 286291.

66 |

Clin. Oral Impl. Res. 20 (Suppl. 4), 2009 / 5966

 c 2009 John Wiley & Sons A/S