Influence of Prosthesis Type and Retention Mechanism on Complications with Fixed Implant-Supported Prostheses: A Systematic Review Applying Multivariate Analyses Christopher Millen, BDS, MFDS, MClinDent, MPros1/ Urs Brägger, DMD, Dr Med Dent2/Julia-Gabriela Wittneben, DMD, Dr Med Dent, MMSc3 Purpose: To identify the influence of fixed prosthesis type on biologic and technical complication rates in the context of screw versus cement retention. Furthermore, a multivariate analysis was conducted to determine which factors, when considered together, influence the complication and failure rates of fixed implant-supported prostheses. Materials and Methods: Electronic searches of MEDLINE (PubMed), EMBASE, and the Cochrane Library were conducted. Selected inclusion and exclusion criteria were used to limit the search. Data were analyzed statistically with simple and multivariate random-effects Poisson regressions. Results: Seventy-three articles qualified for inclusion in the study. Screw-retained prostheses showed a tendency toward and significantly more technical complications than cemented prostheses with single crowns and fixed partial prostheses, respectively. Resin chipping and ceramic veneer chipping had high mean event rates, at 10.04 and 8.95 per 100 years, respectively, for full-arch screwed prostheses. For “all fixed prostheses” (prosthesis type not reported or not known), significantly fewer biologic and technical complications were seen with screw retention. Multivariate analysis revealed a significantly greater incidence of technical complications with cemented prostheses. Full-arch prostheses, cantilevered prostheses, and “all fixed prostheses” had significantly higher complication rates than single crowns. A significantly greater incidence of technical and biologic complications was seen with cemented prostheses. Conclusion: Screwretained fixed partial prostheses demonstrated a significantly higher rate of technical complications and screw-retained full-arch prostheses demonstrated a notably high rate of veneer chipping. When “all fixed prostheses” were considered, significantly higher rates of technical and biologic complications were seen for cement-retained prostheses. Multivariate Poisson regression analysis failed to show a significant difference between screw- and cement-retained prostheses with respect to the incidence of failure but demonstrated a higher rate of technical and biologic complications for cement-retained prostheses. The incidence of technical complications was more dependent upon prosthesis and retention type than prosthesis or abutment material. Int J Oral Maxillofac Implants 2015;30:110–124. doi: 10.11607/jomi.3607 Key words: cement, complication, dental implants, fixed dental prostheses, multivariate analysis, screw
1ITI
Scholar, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland; Clinical Lecturer/Honorary Specialist Registrar in Restorative Dentistry, Edinburgh Dental Institute, University of Edinburgh, Edinburgh, United Kingdom. 2Professor, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland. 3Assistant Professor, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland; Lecturer in Restorative Dentistry and Biomaterials Sciences, Harvard School of Dental Medicine, Cambridge, Massachusetts. Correspondence to: Dr Julia-Gabriela Wittneben, Department of Reconstructive Dentistry and Gerodontology, Freiburgstrasse 7, CH-3010 Bern, Switzerland. Fax: +41-31-632-49-31. Email: [email protected] ©2015 by Quintessence Publishing Co Inc.
F
ixed implant-supported prostheses may be either screw- or cement-retained. The success rate of these prostheses does not seem to be affected by the type of retention.1,2 However, both retention types have relative advantages and disadvantages3–7 and may have an influence on the frequency of technical and biologic complications.8 A complication is defined in the Glossary of Oral and Maxillofacial Implants as an “Unexpected deviation from the normal treatment outcome. It is generally classified as either technical or biological, eg, surgical complication, hemorrhage, damage to the inferior alveolar nerve, infection, delayed wound healing, or lack of osseointegration.”9 Mechanical and technical risks affecting these complications were elaborated upon and further defined by Salvi and Brägger.10
110 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
A recent comprehensive systematic review on the subject of screw- vs cement-retained prostheses1 grouped event rates into cement- versus screwretained single crowns (SCs), fixed partial prostheses (FPPs), and full-arch fixed dental prostheses (FAFDPs). No statistically significant differences were reported for fixed prosthesis survival. Estimated biologic complication rates (bone loss > 2 mm) were found to be higher with cemented prostheses, whereas screw-retained prostheses exhibited more technical complications. Based on their improved retrievability, screw-retained prostheses were given preference. The subject was visited again by this group of authors at the recent ITI Consensus Conference in Bern, Switzerland.2 The complication data in that study were not separated by prosthesis type, but it was similar to the study of Sailer et al1 in that survival rates, technical complication rates, and biologic complication rates of all screw-retained and all cemented prostheses were analyzed. The results demonstrated no significant difference between retention types for survival but showed more technical and biologic complications for cemented prostheses. In addition to the retention type, the particular components, superstructures, and materials are also likely to influence complication rates. The resulting combination of these factors varies throughout the literature, and it is therefore difficult to determine the true reason for the results seen. An accepted method of statistically controlling for these factors in a systematic review is to apply a multivariate analysis. This was not carried out in either of the aforementioned reviews. Because a separate analysis of complications by prosthesis type was not carried out in the previous review,2 the aim of the current review was to identify the influence of prosthesis type on technical and biologic complication rates in the context of screw versus cement retention. Furthermore, the authors used a multivariate analysis to determine which factors (from among retention type, prostheses type, prosthesis material, and abutment material) influence the complication and failure rates of implant-supported fixed prostheses when considered together.
MATERIALS AND METHODS Study Selection
An initial electronic search using medical subject headings and free-text terms related to the review topic was conducted of EMBASE, MEDLINE (via PubMed), and the Cochrane Library. The search included articles published between 2000 and September 2012 and included articles in English, German, and French. This search was based on the following PICO question:
What is the clinical performance of implant-supported reconstructions (including complications and failures) in patients with edentulous sites treated with either screw or cement retention? Well-defined inclusion and exclusion criteria were applied to narrow the search. These criteria dictated that the study had to be prospective, with a mean followup time of 3 years and a minimum of 10 patients. Retrospective studies were included if a follow-up clinical examination had been conducted. The study also had to clearly state the type of retention used (screw or cement) and had to have assessed implant-supported fixed prostheses. Data from patient cohorts used for multiple publications were limited to the most recent version. After the search had been narrowed by the inclusion and exclusion criteria, a hand search was performed of the bibliographies of recently published, relevant review articles. This hand search was not limited by date and included studies published earlier than the year 2000. The full texts of the remaining articles were screened to determine their suitability for inclusion, and data were extracted from each of the qualifying articles during this process. Data extracted included: author, year, type of study (prospective/retrospective), planned number of patients, actual number of patients, mean patient age, age range of patients, study setting (university/private practice), location (anterior/ posterior), fixed prosthesis type, abutment material, prosthesis material, retention type, cement type, implant brand, implant types, and total number of implants. The total exposure time of the prostheses was calculated, and survival of the prostheses was defined as remaining in situ throughout the study period. Data regarding technical complications were also extracted. Technical complications included loss of retention, loosening of the occlusal/abutment screws, loss of screw access fill material, fracture and/or chipping of the veneer, and fracture of the implant/abutment/ framework/screw. Biologic complications were defined as bone loss > 2 mm, peri-implantitis, peri-implant mucositis, general soft tissue complications (including fistulae and swelling), recession, loss of the implant, any esthetic complication, and any other reported complications. The data were grouped and assessed according to the prosthesis type. Namely, this encompassed SCs, FPPs, FAFDPs, and cantilever FDPs (cFDPs). Studies that did not report the prosthesis type or in which identification of the prosthesis type was not possible were grouped together as “all fixed prostheses.” Likewise, studies that did not report on abutment or prosthetic material were grouped as “all materials” and “all abutment materials,” respectively. Individual and summative analyses, as well as a multivariate analysis, were carried out to add to the information already presented by Wittneben et al.2 The International Journal of Oral & Maxillofacial Implants 111
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Initial electronic search: 4,324 articles
Individual selection after applying inclusion/exclusion criteria by two reviewers (abstract search) (reviewer 1): 346; (reviewer 2): 333
Agreement on the selected articles by discussion (reviewers 1 and 2): 302
Hand search (reviewer 3): 19
Abstracts selected for full text review: 321
complications were not considered were excluded from this analysis. The summary analyses for technical and biologic complications summarized the cumulative rates of complications, excluding resin chipping and implant loss, respectively. The estimated event rate per 100 years was calculated by using the observation time of the studies together with the number of prostheses observed (eg, 100 prostheses observed for 1 year each, with only one failure, would have an event rate of 1 per 100 years). Comparisons included differences in event rates per 100 prosthesis-years between cement- and screwretained prostheses, in total and when grouped according to prosthesis type. In addition, a multivariate random-effects Poisson regression was used to compare the incidence rate ratios (IRR) of different groups, with the IRR being the ratio between incidence rates of two parameters. The threshold for significance was set at .05 and P values were calculated using a Wald test.
RESULTS Full review of 321 studies by two reviewers
Data extraction into spreadsheet of 321 studies (reviewers 1 and 3)
Individual selection of the final articles by two reviewers: 73
Fig 1 Flow diagram describing the search design and strategy.
Statistical Analysis
All statistical analysis was performed using Stata 11.2. Random-effects Poisson regression analysis was used to estimate the complication rate and calculated 95% confidence intervals (CIs). For all analyses, the total exposure time per study was used as an offset variable (ln[exposure time]). When no complication occurred, no confidence intervals were calculated. Complication rates of individual studies were calculated by dividing the number of events by the total exposure time. Estimated complication rates and 95% CIs were calculated by assuming a Poisson distributed number of events and by assuming constant event rates. Studies that reported on at least one of the complications examined were included in the overall analysis. Where there were unreported complications, these were assumed to be zero. Studies in which
The titles and abstracts of 4,324 articles (initial search) were screened independently by two authors (JW, UB) to determine their suitability for inclusion in the review (Fig 1). Following discussion, a consensus was reached regarding disputed articles. Three hundred two articles were obtained for screening. An additional 19 articles were obtained from a manual search of the bibliographies of review articles identified within the initial search and recently published relevant reviews. Two authors (JW, CM) independently reviewed the 321 articles. Of these articles, 73 qualified for inclusion in the review.5,11–82 The qualifying articles included 52 prospective cohort studies (71.2%), 13 retrospective studies (17.8%), 2 split-mouth studies, and 6 randomized controlled trials (8.3%). Most studies were carried out in a university setting (63%) (Tables 1 and 2). The types of prostheses studied are summarized in Table 3.
Findings of Simple Poisson Regression Analysis
Technical Complications by Prosthesis Type. The complications “loss of coverage of access hole” and “loosening of occlusal screw” could not be compared statistically, as they were only applicable for screwretained prostheses; however, they were included in the summary analyses. A statistical analysis was performed for the following observations: loss of retention, fracture and/or chipping, loosening of abutment, fracture of abutment, fracture of framework, fracture of implant, screw fracture, resin chipping and/or fracture, and “other.”
112 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
When comparing the summative overall technical complications between cement and screw retention for each prosthesis type, the “resin chipping” category was removed. It could not be analyzed, because it was not reported in some of the cemented prosthesis studies and represented a significant outlier for screwretained complications. Single Crowns. Technical complications with SCs are summarized in Table 4. There were no fractured implants and no episodes of resin chipping associated with SCs. “Loosening of abutment” showed a statistically significantly higher rate in screw-retained prostheses (P = .003), whereas the opposite was true for “framework fracture” (P < .001). Complication rates per 100 years did not exceed 1, except for “loosening of abutment” (2.07 [0.65 to 6.58] per 100 years) and “fracture of framework” (2.69 [1.59 to 4.54] per 100 years). Overall, there was no difference between retention types; however, there was a tendency toward more complications with screw-retained prostheses (P = .071). Fixed Partial Prostheses . Technical complications with FPPs are summarized in Table 5. There were no statistically significant differences with any of the individual complications. However, the summary of all complications showed statistically significantly more events in the screw-retained group (P = .009). Full-Arch FDPs. Technical complications with FAFDPs are summarized in Table 6. Only one included study reported on cemented FAFDPs; however, it did not report on any complications. Therefore, no statistical comparison was possible. Resin chipping and ceramic veneer chipping had high event rates, at 10.04 (4.48 to 22.48) and 8.95 (1.25 to 64.13) per 100 years, respectively. Cantilever FDPs. Technical complications with cFDPs are summarized in Table 7. The number of included studies reporting on complications for cFDPs was low, with only 1 for cemented and 2 for screw-retained cFDPs. Because of the small numbers of studies, complications were rarely considered, and as such statistical comparisons were not possible. Rates of 10.33 and 11.59 for ceramic veneer chipping and abutment loosening, respectively, were high in comparison to the calculated rates for other types of prostheses. A large difference in overall complication rates per 100 years was reported between screw-retained (17.93) and cemented (0.59) cFDPs. All Fixed Prostheses. Technical complications in the group “all fixed prostheses” are summarized in Table 8. There were statistically significantly higher rates of retention loss and abutment loosening in the cemented groups (P < .001), which contributed to a greater number of complications with cemented prostheses (P < .001).
Table 1 Design of Included Studies Study type
No. of studies
%
Prospective cohort
52
71.2
Retrospective cohort
13
17.8
2
2.7
Split mouth Randomized controlled Total
6
8.3
73
100
Table 2 Settings of Included Studies No. of studies
%
Private practice
Study setting
13
17.8
University
46
63
Specialist clinic
6
8.2
Multicenter
6
8.2
Not reported
2
Total
2.8
73
100
Table 3 Types of Implant-Supported Prostheses Analyzed in Included Studies Reported prosthesis type
No. of prostheses
%
Single crown
1,720
29.4
Fixed partial
979
16.7
Full-arch FDP
928
15.8
Cantilever FDP
61
1.0
All fixed prostheses
2,170
36.2
Total
5,858
100
Biologic Complications by Prosthesis Type. The same method used for technical complications— excluding studies from analysis that did not report on individual complications—was used to analyze biologic complications. The summary assessments for each prosthesis type did not include any data on implant loss. Single Crowns. Biologic complications with SCs are summarized in Table 9. Only the category “any esthetic complications” showed a statistically significant result, with a higher complication rate for screw-retained crowns (P = .01). Summary results showed no difference in biologic complication rates between the two retention types. Fixed Partial Prostheses. Biologic complications with FPPs are summarized in Table 10. Overall, cemented FDPs in the included studies reported no biologic complications, and screw-retained prostheses showed a complication rate of 4.18 (2.11 to 8.28) per 100 years. The International Journal of Oral & Maxillofacial Implants 113
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 4 Technical Complications with SCs Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement
22 6 0 6 22
1,193 309 0 309 1,126
6,155 1,196 – 1,196 5,770
23 10 – 1 23
0.55 (0.25–1.23) 0.47 (0.12–1.79) – (–) 0.08 (0.01–0.59) 0.41 (0.26–0.65)
Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
7 0 8 22 6 22 6 2 7 22 6 22 7 1 6 20 7 24 8
355 0 373 1,126 309 1,126 309 109 355 1,126 309 1,126 355 12 309 958 327 1,246 373
1,426 – 1,606 5,770 1,196 5,770 1,196 521 1,426 5,770 1,196 5,770 1,426 36 1,196 4,611 1,376 6,485 1,606
4 – 13 18 23 0 1 14 1 0 0 0 4 0 0 33 7 111 64
0.28 (0.10–0.79) – (–) 0.90 (0.20–4.09) 0.33 (0.15–0.71) 2.07 (0.65–6.58) 0 (–) 0.08 (0.01–0.59) 2.69 (1.59–4.54) 0.07 (0.01–0.50) 0 (–) 0 (–) 0 (–) 0.31 (0.07–1.32) 0 (–) 0 (–) 0.60 (0.19–1.86) 0.30 (0.08–1.14) 1.91 (1.25–2.94) 3.12 (1.79–5.42)
P .834 N/A .523 N/A .003 > .999 < .001 N/A > .999 N/A .145 .071
*Estimated complication rate per 100 years (95% confidence interval).
Table 5 Technical Complications with FPPs Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications 95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
3 9 0 9 3 11 0 10 3 9 3 8 1 10 3 9 3 11 1 9 3 10 3 11
61 460 0 460 61 649 0 529 61 486 61 340 16 529 61 486 61 649 16 486 61 606 61 649
318 2,697 – 2,697 318 4,299 – 3,099 318 2,884 318 1,497 48 3,099 318 2,884 318 4,299 48 2,884 318 4,084 318 4,299
4 16 – 21 1 98 – 34 1 30 0 1 0 1 0 8 0 11 0 6 0 5 6 225
0.94 (0.09–9.72) 0.37 (0.09–1.55) – (–) 0.46 (0.05–3.95) 0.30 (0.02–4.16) 2.80 (1.20–6.57) – (–) 1.83 (0.47–7.02) 0.29 (0.02–3.44) 1.02 (0.40–2.57) 0 (–) 0.07 (0.01–0.47) 0 (–) 0.03 (0.00–0.23) 0 (–) 0.43 (0.11–1.73) 0 (–) 0.21 (0.08–0.53) 0 (–) 0.46 (0.01–14.62) 0 (–) 0.07 (0.01–0.58) 1.19 (0.36–3.86) 6.60 (3.71–11.72)
P .500 N/A .112 N/A .351 > .999 > .999 > .999 > .999 > .999 > .999 .009
*Estimated complication rate per 100 years (95% confidence interval).
114 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 6 Technical Complications with FAFDPs Complication
No. of No. of Exposure No. of Retention studies prostheses time (y) complications
Complication rate* (95% CI)
Loss of retention
Screw
16
851
6,508
30
1.13 (0.26–4.87)
Exposure of access hole
Screw
14
778
6,252
106
1.54 (0.52–4.56)
Fracture and/or chipping
Screw
14
778
6,252
60
8.95 (1.25–64.13)
Loosening of occlusal screw
Screw
16
851
6,508
52
1.75 (0.63–4.87)
Loosening of abutment
Screw
16
851
6,508
21
0.74 (0.10–5.33)
Fracture of abutment
Screw
15
822
6,384
14
0.16 (0.05–0.52)
Fracture of framework
Screw
15
822
6,384
55
0.61 (0.19–1.99)
Fracture of implant
Screw
16
851
6,508
3
0.15 (0.00–4.35)
Screw fracture
Screw
16
851
6,508
18
Resin chipping and/or fracture
Screw
15
822
6,384
533
0.52 (0.15–1.88)
Other
Screw
16
851
6,508
231
5.83 (2.77–12.27)
Summary of all complications except “resin chipping and/or fracture”
Screw
16
851
6,508
590
19.44 (11.08–34.09)
10.04 (4.48–22.48)
*Estimated complication rate per 100 years (95% confidence interval).
Table 7 Technical Complications with cFDPs Complication
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI)
Loss of retention
Cement
1
15
169.5
0
0.00
Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
1 0 1 1 2 0 2 1 1 1 1 0 2 1 1 1 2 0 1 1 1 1 2
23 0 23 15 46 0 46 15 23 15 23 0 46 15 23 15 46 0 23 15 23 15 46
69.0 – 69.0 169.5 184.0 – 184.0 169.5 69.0 169.5 69.0 – 184.0 169.5 69.0 169.5 184.0 – 69.0 169.5 69.0 169.5 184.0
0 – 0 0 19 – 5 1 8 0 0 – 0 0 0 0 1 – 0 0 0 1 33
0.00 N/A 0.00 0.00 10.33 N/A 2.72 0.59 11.59 0.00 0.00 N/A 0.00 0.00 0.00 0.00 0.54 N/A 0.00 0.00 0.00 0.59 17.93
Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
*Estimated complication rate per 100 years (95% confidence interval).
Full-Arch FDPs. Biologic complications with FAFDPs are summarized in Table 11. As was seen with technical complications, there were no complications reported for cemented FAFDPs. Bone loss, implant loss, and “other” represent the most frequently occurring biologic complications for FAFDPs. The overall biologic
complication rate for screw-retained FAFDPs was 7.61 (4.55 to 12.72) per 100 years. Cantilever FDPs. Only two studies reported on biologic complications for cFDPs (Table 12). The most commonly reported complication was peri-implantitis, with an event rate of 5.31 per 100 years. The International Journal of Oral & Maxillofacial Implants 115
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 8 Technical Complications for “All Fixed Prostheses” Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
6 5 0 3 7 6 0 6 7 5 7 5 0 6 7 6 7 6 0 5 7 6 7 8
746 1,098 0 964 756 1,173 0 1,224 756 1,117 756 1,117 0 1,224 756 1,224 756 1,224 0 1,117 756 1,173 756 1,307
3,751 4,932 – 4,530 3,805 5,267 – 5,634 3,805 5,313 3,805 5,313 – 5,634 3,805 5,634 3,805 5,634 – 5,313 3,805 5,481 3,805 5,883
68 21 – 3 6 31 – 97 66 3 4 4 – 2 2 0 10 13 – 0 0 0 156 174
20.49 (3.56–117.89) 1.72 (0.33–8.89) – (–) 0.23 (0.01–4.90) 1.03 (0.22–4.89) 1.58 (0.48–5.20) – (–) 3.82 (1.53–9.51) 1.79 (0.70–4.62) 0.06 (0.01–0.27) 0.06 (0.01–0.38) 0.03 (0.00–0.26) – (–) 0.10 (0.01–1.85) 0.13 (0.01–2.44) 0 (–) 0.22 (0.03–1.52) 0.14 (0.02–0.86) – (–) 0 (–) 0 (–) 0 (–) 8.79 (5.20–14.88) 4.70 (2.87–7.69)
P < .001 N/A .507 N/A < .001 .470 N/A .999 .286 N/A N/A < .001
*Estimated complication rate per 100 years (95% confidence interval).
Table 9 Biologic Complications with SCs Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
20 7 19 6 20 7 18 6 18 7 20 7 20 6 20 6 22 8
1,111 327 1,022 309 1,044 327 967 309 882 327 1,075 355 1,117 309 1,142 309 1,195 373
6,028 1,376 4,590 1,196 5,633 1,376 4,040 1,196 3,946 1,376 5,565 1,426 5,500 1,196 5,960 1,196 6,280 1,606
34 1 15 0 20 2 59 2 6 1 15 3 10 14 43 14 187 34
0.41 (0.19–0.88) 0.08 (0.01–0.94) 0.24 (0.08–0.71) 0 (–) 0.41 (0.18–0.93) 0.24 (0.04–1.49) 1.71 (0.53–5.57) 1.16 (0.14–9.90) 0.19 (0.05–0.69) 0.12 (0.01–1.55) 0.27 (0.16–0.47) 0.22 (0.07–0.72) 0.19 (0.06–0.58) 1.09 (0.30–4.02) 1.22 (0.41–3.65) 0.93 (0.28–3.11) 3.40 (1.98–5.82) 4.46 (2.07–9.58)
P .213 .999 .567 .687 .704 .736 .010 .698 .425
*Estimated complication rate per 100 years (95% confidence interval).
All Fixed Prostheses. Biologic complications with “all fixed prostheses” are summarized in Table 13. There were statistically significantly more overall complications with cemented prostheses (P < .001). No
individual complication reached significance, but there was a tendency toward more implant loss for cemented prostheses (P = .074).
116 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 10
Biologic Complications with FPPs
Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
2 8 2 8 2 8 1 8 1 7 2 10 2 9 2 9 2 10
25 435 25 461 25 461 16 435 16 315 25 624 25 581 25 581 25 624
150 2,572 150 2,759 150 2,759 48 2,572 48 1,372 150 4,174 150 3,959 150 3,959 150 4,174
0 28 0 10 0 11 0 59 0 0 0 33 0 9 0 32 0 149
0 (–) 1.35 (0.25–7.27) 0 (–) 0.42 (0.13–1.34) 0 (–) 0.38 (0.16–0.90) 0 (–) 2.04 (0.83–5.02) 0 (–) 0 (–) 0 (–) 0.76 (0.34–1.70) 0 (–) 0.17 (0.04–0.77) 0 (–) 0.98 (0.20–4.84) 0 (–) 4.18 (2.11–8.28)
P > .999 > .999 > .999 .99 N/A > .999 > .999 > .999 > .999
*Estimated complication rate per 100 years (95% confidence interval).
Table 11 Biologic Complications with FAFDPs Complication
No. of No. of Exposure No. of Retention studies prostheses time (y) complications
Complication rate* (95% CI)
Bone loss (> 2 mm) Peri-implantitis
Screw Screw
13 11
742 651
6,121 5,811
356 0
3.28 (1.14–9.44) 0 (–)
Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
Screw Screw Screw Screw Screw Screw Screw
11 12 11 13 11 12 14
651 680 651 729 651 713 758
5,811 5,935 5,811 6,045 5,811 5,997 6,169
23 100 0 70 1 161 641
0.62 (0.06–6.97) 1.61 (0.56–4.62) 0 (–) 2.21 (0.73–6.71) 0.03 (0.00–0.51) 2.94 (1.36–6.33) 7.61 (4.55–12.72)
*Estimated complication rate per 100 years (95% confidence interval).
Outcome of Multivariate Poisson Regression Analysis
Prosthesis Failure The results and incidence rate ratios for this analysis are presented in Table 14. There was no statistically significant difference between the failure rates for cemented and screw-retained prostheses. Technical Complications. The results are presented in Table 15. There was a statistically significantly greater incidence of technical complications in cemented compared to screw-retained prostheses (P = .01). FAFDPs (P < .001), cFDPs (P < .001), and “all fixed prostheses” (P = .003) had significantly higher complication rates than SCs. The complication rates of acrylic resin, ceramic, and “all materials” did not dif-
fer significantly from that of porcelain-fused-to-metal (PFM) prostheses. The complication rates of gold and ceramic abutment materials did not differ significantly from that of titanium abutments. Titanium abutments showed a tendency toward higher complication rates than the “all abutment materials” group (P = .051). Biologic Complications. The results are presented in Table 1. There was a statistically significantly greater incidence of biologic complications in cemented than in screw-retained prostheses (P = .001, IRR 1.73). “All fixed prostheses” had a statistically significantly higher complication rate than SCs (P = .032). Cantilever prostheses showed a tendency to have a higher complication rate than SCs (P = .051). The International Journal of Oral & Maxillofacial Implants 117
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 12 Biologic Complications with cFDPs No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI)
Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
1 0 1 0 1 0 0 0 0 0 1 1 1 0 1 0 1 1
15 0 15 0 15 0 0 0 0 0 15 23 15 0 15 0 15 23
169.5 – 169.5 – 169.5 – – – – – 169.5 115.0 169.5 – 169.5 – 169.5 115.0
0 – 9 – 0 – – – – – 2 1 0 – 0 – 9 0
0.00 N/A 5.31 N/A 0.00 N/A N/A N/A N/A N/A 1.18 0.87 0.00 N/A 0.00 N/A 5.31 0.00
*Estimated complication rate per 100 years (95% confidence interval).
Table 13 Biologic Complications for “All Fixed Prostheses” Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
6 5 6 5 6 5 6 4 6 4 6 6 6 4 6 5 6 6
629 1,128 629 1,128 629 1,128 629 1,072 629 1,072 629 1,235 629 1,072 629 1,128 629 1,235
3,149 5,345 3,149 5,345 3,149 5,345 3,149 5,177 3,149 5,177 3,149 5,666 3,149 5,177 3,149 5,345 3,149 5,666
5 85 22 38 35 0 1 6 0 0 40 76 7 0 2 0 72 129
P
2.89 (0.52–16.09) .50 4.27 (1.11–16.39) 1.74 (0.38–7.94) .18 1.19 (0.28–5.10) 1.27 (0.04–44.33) > .999 0 (–) 0.06 (0.01–0.72) .27 0.21 (0.04–1.18) 0 (–) N/A 0 (–) 2.29 (0.90–5.79) .07 1.54 (0.65–3.63) 0.30 (0.02–4.77) > .999 0 (–) 0.10 (0.01–1.07) .99 0 (–) 12.89 (3.45–48.16) < .001 5.47 (1.53–19.59)
*Estimated complication rate per 100 years (95% confidence interval).
DISCUSSION Restoring partially edentulous patients with the use of dental implants has been reported as a successful treatment approach, with high success and survival rates for both implants and implant-supported prostheses.83,84 Many factors, including the clinician’s preference,
influence the choice of retention. It is plausible to restore a single edentulous gap with an implant-supported SC, but to rehabilitate an extended edentulous gap, the use of several implants might be considered, depending on the planned prosthetic outcome. In this situation, a fixed prosthesis may consist of solely implant-supported FPP, a cFDP, an FAFDP, or several implant-supported SCs.
118 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 14 Multivariate Analysis of Failure Estimated IRR Retention Cement vs screw Type of prosthesis FPD+cantilever vs SC FAFDP vs SC All fixed prostheses vs SC Prosthesis material Acrylic vs PFM Ceramic vs PFM All materials vs PFM Abutment material Gold vs titanium Ceramic vs titanium All abutment materials vs titanium
Table 15 Multivariate Analysis of Technical Complications 95% CI
P
0.92
0.50–1.69
.794
1.91 1.63 0.41
0.86–4.27 0.49–5.39 0.14–1.17
.113 .424 .096
0.60 1.34 3.45
0.16–2.29 0.38–4.74 1.44–8.26
.457 .654 .005
2.24 1.58 0.58
0.71–7.05 0.43–5.83 0.19–1.78
.167 .491 .341
The purpose of this investigation was to identify the influence of fixed prosthesis and retention type on technical and biologic complication rates. The European Association of Osseointegration (EAO) and the International Team for Implantology (ITI) recently published consensus statements and clinical recommendations regarding the question of whether screw or cement retention is preferred for fixed prostheses.85,86 These two sets of statements and recommendations complement one another and indicate a preference for screw retention based on retrievability and a higher frequency of biologic complications with cemented prostheses. The complication rates of each prosthesis type in the current review revealed a variety of different outcomes. There was a tendency toward more technical complications with screw-retained SCs, without statistical significance. More specifically, there were significantly more instances of abutment loosening with screw-retained crowns, which was in accordance with the results of the systematic review of Sailer et al.1 More framework fractures occurred with cemented SCs; however, this could not be compared to Sailer et al, as data on framework fracture were not presented in their published article. No statistically significant differences were identified for biologic complications relative to retention type for single implant–supported crowns, whereas in the previous review from Sailer et al,1 more soft tissue complications were documented with screw-retained SCs. This difference may have resulted because of esthetic complications that, while not considered by Sailer et al, contributed to the overall biologic complications in the current review. Esthetic complications may be of particular importance to patients.87
Retention Cement vs screw Type of prosthesis FPP vs SC FAFDP vs SC cFDP vs SC All fixed prostheses vs SC Prosthesis material Acrylic vs PFM Ceramic vs PFM All materials vs PFM Abutment material Gold vs titanium Ceramic vs titanium All abutment materials vs titanium
Estimated IRR
95% CI
1.36
1.08–1.72
1.58 9.80
0.88–2.84 .127 3.64–26.38
1ITI
Scholar, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland; Clinical Lecturer/Honorary Specialist Registrar in Restorative Dentistry, Edinburgh Dental Institute, University of Edinburgh, Edinburgh, United Kingdom. 2Professor, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland. 3Assistant Professor, Department of Reconstructive Dentistry and Gerodontology, School of Dental Medicine, University of Bern, Switzerland; Lecturer in Restorative Dentistry and Biomaterials Sciences, Harvard School of Dental Medicine, Cambridge, Massachusetts. Correspondence to: Dr Julia-Gabriela Wittneben, Department of Reconstructive Dentistry and Gerodontology, Freiburgstrasse 7, CH-3010 Bern, Switzerland. Fax: +41-31-632-49-31. Email: [email protected] ©2015 by Quintessence Publishing Co Inc.
F
ixed implant-supported prostheses may be either screw- or cement-retained. The success rate of these prostheses does not seem to be affected by the type of retention.1,2 However, both retention types have relative advantages and disadvantages3–7 and may have an influence on the frequency of technical and biologic complications.8 A complication is defined in the Glossary of Oral and Maxillofacial Implants as an “Unexpected deviation from the normal treatment outcome. It is generally classified as either technical or biological, eg, surgical complication, hemorrhage, damage to the inferior alveolar nerve, infection, delayed wound healing, or lack of osseointegration.”9 Mechanical and technical risks affecting these complications were elaborated upon and further defined by Salvi and Brägger.10
110 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
A recent comprehensive systematic review on the subject of screw- vs cement-retained prostheses1 grouped event rates into cement- versus screwretained single crowns (SCs), fixed partial prostheses (FPPs), and full-arch fixed dental prostheses (FAFDPs). No statistically significant differences were reported for fixed prosthesis survival. Estimated biologic complication rates (bone loss > 2 mm) were found to be higher with cemented prostheses, whereas screw-retained prostheses exhibited more technical complications. Based on their improved retrievability, screw-retained prostheses were given preference. The subject was visited again by this group of authors at the recent ITI Consensus Conference in Bern, Switzerland.2 The complication data in that study were not separated by prosthesis type, but it was similar to the study of Sailer et al1 in that survival rates, technical complication rates, and biologic complication rates of all screw-retained and all cemented prostheses were analyzed. The results demonstrated no significant difference between retention types for survival but showed more technical and biologic complications for cemented prostheses. In addition to the retention type, the particular components, superstructures, and materials are also likely to influence complication rates. The resulting combination of these factors varies throughout the literature, and it is therefore difficult to determine the true reason for the results seen. An accepted method of statistically controlling for these factors in a systematic review is to apply a multivariate analysis. This was not carried out in either of the aforementioned reviews. Because a separate analysis of complications by prosthesis type was not carried out in the previous review,2 the aim of the current review was to identify the influence of prosthesis type on technical and biologic complication rates in the context of screw versus cement retention. Furthermore, the authors used a multivariate analysis to determine which factors (from among retention type, prostheses type, prosthesis material, and abutment material) influence the complication and failure rates of implant-supported fixed prostheses when considered together.
MATERIALS AND METHODS Study Selection
An initial electronic search using medical subject headings and free-text terms related to the review topic was conducted of EMBASE, MEDLINE (via PubMed), and the Cochrane Library. The search included articles published between 2000 and September 2012 and included articles in English, German, and French. This search was based on the following PICO question:
What is the clinical performance of implant-supported reconstructions (including complications and failures) in patients with edentulous sites treated with either screw or cement retention? Well-defined inclusion and exclusion criteria were applied to narrow the search. These criteria dictated that the study had to be prospective, with a mean followup time of 3 years and a minimum of 10 patients. Retrospective studies were included if a follow-up clinical examination had been conducted. The study also had to clearly state the type of retention used (screw or cement) and had to have assessed implant-supported fixed prostheses. Data from patient cohorts used for multiple publications were limited to the most recent version. After the search had been narrowed by the inclusion and exclusion criteria, a hand search was performed of the bibliographies of recently published, relevant review articles. This hand search was not limited by date and included studies published earlier than the year 2000. The full texts of the remaining articles were screened to determine their suitability for inclusion, and data were extracted from each of the qualifying articles during this process. Data extracted included: author, year, type of study (prospective/retrospective), planned number of patients, actual number of patients, mean patient age, age range of patients, study setting (university/private practice), location (anterior/ posterior), fixed prosthesis type, abutment material, prosthesis material, retention type, cement type, implant brand, implant types, and total number of implants. The total exposure time of the prostheses was calculated, and survival of the prostheses was defined as remaining in situ throughout the study period. Data regarding technical complications were also extracted. Technical complications included loss of retention, loosening of the occlusal/abutment screws, loss of screw access fill material, fracture and/or chipping of the veneer, and fracture of the implant/abutment/ framework/screw. Biologic complications were defined as bone loss > 2 mm, peri-implantitis, peri-implant mucositis, general soft tissue complications (including fistulae and swelling), recession, loss of the implant, any esthetic complication, and any other reported complications. The data were grouped and assessed according to the prosthesis type. Namely, this encompassed SCs, FPPs, FAFDPs, and cantilever FDPs (cFDPs). Studies that did not report the prosthesis type or in which identification of the prosthesis type was not possible were grouped together as “all fixed prostheses.” Likewise, studies that did not report on abutment or prosthetic material were grouped as “all materials” and “all abutment materials,” respectively. Individual and summative analyses, as well as a multivariate analysis, were carried out to add to the information already presented by Wittneben et al.2 The International Journal of Oral & Maxillofacial Implants 111
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Initial electronic search: 4,324 articles
Individual selection after applying inclusion/exclusion criteria by two reviewers (abstract search) (reviewer 1): 346; (reviewer 2): 333
Agreement on the selected articles by discussion (reviewers 1 and 2): 302
Hand search (reviewer 3): 19
Abstracts selected for full text review: 321
complications were not considered were excluded from this analysis. The summary analyses for technical and biologic complications summarized the cumulative rates of complications, excluding resin chipping and implant loss, respectively. The estimated event rate per 100 years was calculated by using the observation time of the studies together with the number of prostheses observed (eg, 100 prostheses observed for 1 year each, with only one failure, would have an event rate of 1 per 100 years). Comparisons included differences in event rates per 100 prosthesis-years between cement- and screwretained prostheses, in total and when grouped according to prosthesis type. In addition, a multivariate random-effects Poisson regression was used to compare the incidence rate ratios (IRR) of different groups, with the IRR being the ratio between incidence rates of two parameters. The threshold for significance was set at .05 and P values were calculated using a Wald test.
RESULTS Full review of 321 studies by two reviewers
Data extraction into spreadsheet of 321 studies (reviewers 1 and 3)
Individual selection of the final articles by two reviewers: 73
Fig 1 Flow diagram describing the search design and strategy.
Statistical Analysis
All statistical analysis was performed using Stata 11.2. Random-effects Poisson regression analysis was used to estimate the complication rate and calculated 95% confidence intervals (CIs). For all analyses, the total exposure time per study was used as an offset variable (ln[exposure time]). When no complication occurred, no confidence intervals were calculated. Complication rates of individual studies were calculated by dividing the number of events by the total exposure time. Estimated complication rates and 95% CIs were calculated by assuming a Poisson distributed number of events and by assuming constant event rates. Studies that reported on at least one of the complications examined were included in the overall analysis. Where there were unreported complications, these were assumed to be zero. Studies in which
The titles and abstracts of 4,324 articles (initial search) were screened independently by two authors (JW, UB) to determine their suitability for inclusion in the review (Fig 1). Following discussion, a consensus was reached regarding disputed articles. Three hundred two articles were obtained for screening. An additional 19 articles were obtained from a manual search of the bibliographies of review articles identified within the initial search and recently published relevant reviews. Two authors (JW, CM) independently reviewed the 321 articles. Of these articles, 73 qualified for inclusion in the review.5,11–82 The qualifying articles included 52 prospective cohort studies (71.2%), 13 retrospective studies (17.8%), 2 split-mouth studies, and 6 randomized controlled trials (8.3%). Most studies were carried out in a university setting (63%) (Tables 1 and 2). The types of prostheses studied are summarized in Table 3.
Findings of Simple Poisson Regression Analysis
Technical Complications by Prosthesis Type. The complications “loss of coverage of access hole” and “loosening of occlusal screw” could not be compared statistically, as they were only applicable for screwretained prostheses; however, they were included in the summary analyses. A statistical analysis was performed for the following observations: loss of retention, fracture and/or chipping, loosening of abutment, fracture of abutment, fracture of framework, fracture of implant, screw fracture, resin chipping and/or fracture, and “other.”
112 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
When comparing the summative overall technical complications between cement and screw retention for each prosthesis type, the “resin chipping” category was removed. It could not be analyzed, because it was not reported in some of the cemented prosthesis studies and represented a significant outlier for screwretained complications. Single Crowns. Technical complications with SCs are summarized in Table 4. There were no fractured implants and no episodes of resin chipping associated with SCs. “Loosening of abutment” showed a statistically significantly higher rate in screw-retained prostheses (P = .003), whereas the opposite was true for “framework fracture” (P < .001). Complication rates per 100 years did not exceed 1, except for “loosening of abutment” (2.07 [0.65 to 6.58] per 100 years) and “fracture of framework” (2.69 [1.59 to 4.54] per 100 years). Overall, there was no difference between retention types; however, there was a tendency toward more complications with screw-retained prostheses (P = .071). Fixed Partial Prostheses . Technical complications with FPPs are summarized in Table 5. There were no statistically significant differences with any of the individual complications. However, the summary of all complications showed statistically significantly more events in the screw-retained group (P = .009). Full-Arch FDPs. Technical complications with FAFDPs are summarized in Table 6. Only one included study reported on cemented FAFDPs; however, it did not report on any complications. Therefore, no statistical comparison was possible. Resin chipping and ceramic veneer chipping had high event rates, at 10.04 (4.48 to 22.48) and 8.95 (1.25 to 64.13) per 100 years, respectively. Cantilever FDPs. Technical complications with cFDPs are summarized in Table 7. The number of included studies reporting on complications for cFDPs was low, with only 1 for cemented and 2 for screw-retained cFDPs. Because of the small numbers of studies, complications were rarely considered, and as such statistical comparisons were not possible. Rates of 10.33 and 11.59 for ceramic veneer chipping and abutment loosening, respectively, were high in comparison to the calculated rates for other types of prostheses. A large difference in overall complication rates per 100 years was reported between screw-retained (17.93) and cemented (0.59) cFDPs. All Fixed Prostheses. Technical complications in the group “all fixed prostheses” are summarized in Table 8. There were statistically significantly higher rates of retention loss and abutment loosening in the cemented groups (P < .001), which contributed to a greater number of complications with cemented prostheses (P < .001).
Table 1 Design of Included Studies Study type
No. of studies
%
Prospective cohort
52
71.2
Retrospective cohort
13
17.8
2
2.7
Split mouth Randomized controlled Total
6
8.3
73
100
Table 2 Settings of Included Studies No. of studies
%
Private practice
Study setting
13
17.8
University
46
63
Specialist clinic
6
8.2
Multicenter
6
8.2
Not reported
2
Total
2.8
73
100
Table 3 Types of Implant-Supported Prostheses Analyzed in Included Studies Reported prosthesis type
No. of prostheses
%
Single crown
1,720
29.4
Fixed partial
979
16.7
Full-arch FDP
928
15.8
Cantilever FDP
61
1.0
All fixed prostheses
2,170
36.2
Total
5,858
100
Biologic Complications by Prosthesis Type. The same method used for technical complications— excluding studies from analysis that did not report on individual complications—was used to analyze biologic complications. The summary assessments for each prosthesis type did not include any data on implant loss. Single Crowns. Biologic complications with SCs are summarized in Table 9. Only the category “any esthetic complications” showed a statistically significant result, with a higher complication rate for screw-retained crowns (P = .01). Summary results showed no difference in biologic complication rates between the two retention types. Fixed Partial Prostheses. Biologic complications with FPPs are summarized in Table 10. Overall, cemented FDPs in the included studies reported no biologic complications, and screw-retained prostheses showed a complication rate of 4.18 (2.11 to 8.28) per 100 years. The International Journal of Oral & Maxillofacial Implants 113
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 4 Technical Complications with SCs Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement
22 6 0 6 22
1,193 309 0 309 1,126
6,155 1,196 – 1,196 5,770
23 10 – 1 23
0.55 (0.25–1.23) 0.47 (0.12–1.79) – (–) 0.08 (0.01–0.59) 0.41 (0.26–0.65)
Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
7 0 8 22 6 22 6 2 7 22 6 22 7 1 6 20 7 24 8
355 0 373 1,126 309 1,126 309 109 355 1,126 309 1,126 355 12 309 958 327 1,246 373
1,426 – 1,606 5,770 1,196 5,770 1,196 521 1,426 5,770 1,196 5,770 1,426 36 1,196 4,611 1,376 6,485 1,606
4 – 13 18 23 0 1 14 1 0 0 0 4 0 0 33 7 111 64
0.28 (0.10–0.79) – (–) 0.90 (0.20–4.09) 0.33 (0.15–0.71) 2.07 (0.65–6.58) 0 (–) 0.08 (0.01–0.59) 2.69 (1.59–4.54) 0.07 (0.01–0.50) 0 (–) 0 (–) 0 (–) 0.31 (0.07–1.32) 0 (–) 0 (–) 0.60 (0.19–1.86) 0.30 (0.08–1.14) 1.91 (1.25–2.94) 3.12 (1.79–5.42)
P .834 N/A .523 N/A .003 > .999 < .001 N/A > .999 N/A .145 .071
*Estimated complication rate per 100 years (95% confidence interval).
Table 5 Technical Complications with FPPs Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications 95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
3 9 0 9 3 11 0 10 3 9 3 8 1 10 3 9 3 11 1 9 3 10 3 11
61 460 0 460 61 649 0 529 61 486 61 340 16 529 61 486 61 649 16 486 61 606 61 649
318 2,697 – 2,697 318 4,299 – 3,099 318 2,884 318 1,497 48 3,099 318 2,884 318 4,299 48 2,884 318 4,084 318 4,299
4 16 – 21 1 98 – 34 1 30 0 1 0 1 0 8 0 11 0 6 0 5 6 225
0.94 (0.09–9.72) 0.37 (0.09–1.55) – (–) 0.46 (0.05–3.95) 0.30 (0.02–4.16) 2.80 (1.20–6.57) – (–) 1.83 (0.47–7.02) 0.29 (0.02–3.44) 1.02 (0.40–2.57) 0 (–) 0.07 (0.01–0.47) 0 (–) 0.03 (0.00–0.23) 0 (–) 0.43 (0.11–1.73) 0 (–) 0.21 (0.08–0.53) 0 (–) 0.46 (0.01–14.62) 0 (–) 0.07 (0.01–0.58) 1.19 (0.36–3.86) 6.60 (3.71–11.72)
P .500 N/A .112 N/A .351 > .999 > .999 > .999 > .999 > .999 > .999 .009
*Estimated complication rate per 100 years (95% confidence interval).
114 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 6 Technical Complications with FAFDPs Complication
No. of No. of Exposure No. of Retention studies prostheses time (y) complications
Complication rate* (95% CI)
Loss of retention
Screw
16
851
6,508
30
1.13 (0.26–4.87)
Exposure of access hole
Screw
14
778
6,252
106
1.54 (0.52–4.56)
Fracture and/or chipping
Screw
14
778
6,252
60
8.95 (1.25–64.13)
Loosening of occlusal screw
Screw
16
851
6,508
52
1.75 (0.63–4.87)
Loosening of abutment
Screw
16
851
6,508
21
0.74 (0.10–5.33)
Fracture of abutment
Screw
15
822
6,384
14
0.16 (0.05–0.52)
Fracture of framework
Screw
15
822
6,384
55
0.61 (0.19–1.99)
Fracture of implant
Screw
16
851
6,508
3
0.15 (0.00–4.35)
Screw fracture
Screw
16
851
6,508
18
Resin chipping and/or fracture
Screw
15
822
6,384
533
0.52 (0.15–1.88)
Other
Screw
16
851
6,508
231
5.83 (2.77–12.27)
Summary of all complications except “resin chipping and/or fracture”
Screw
16
851
6,508
590
19.44 (11.08–34.09)
10.04 (4.48–22.48)
*Estimated complication rate per 100 years (95% confidence interval).
Table 7 Technical Complications with cFDPs Complication
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI)
Loss of retention
Cement
1
15
169.5
0
0.00
Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
1 0 1 1 2 0 2 1 1 1 1 0 2 1 1 1 2 0 1 1 1 1 2
23 0 23 15 46 0 46 15 23 15 23 0 46 15 23 15 46 0 23 15 23 15 46
69.0 – 69.0 169.5 184.0 – 184.0 169.5 69.0 169.5 69.0 – 184.0 169.5 69.0 169.5 184.0 – 69.0 169.5 69.0 169.5 184.0
0 – 0 0 19 – 5 1 8 0 0 – 0 0 0 0 1 – 0 0 0 1 33
0.00 N/A 0.00 0.00 10.33 N/A 2.72 0.59 11.59 0.00 0.00 N/A 0.00 0.00 0.00 0.00 0.54 N/A 0.00 0.00 0.00 0.59 17.93
Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
*Estimated complication rate per 100 years (95% confidence interval).
Full-Arch FDPs. Biologic complications with FAFDPs are summarized in Table 11. As was seen with technical complications, there were no complications reported for cemented FAFDPs. Bone loss, implant loss, and “other” represent the most frequently occurring biologic complications for FAFDPs. The overall biologic
complication rate for screw-retained FAFDPs was 7.61 (4.55 to 12.72) per 100 years. Cantilever FDPs. Only two studies reported on biologic complications for cFDPs (Table 12). The most commonly reported complication was peri-implantitis, with an event rate of 5.31 per 100 years. The International Journal of Oral & Maxillofacial Implants 115
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 8 Technical Complications for “All Fixed Prostheses” Complication Loss of retention Exposure of access hole Fracture and/or chipping Loosening of occlusal screw Loosening of abutment Fracture of abutment Fracture of framework Fracture of implant Screw fracture Resin chipping and/or fracture Other Summary of all complications except “resin chipping and/or fracture”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
6 5 0 3 7 6 0 6 7 5 7 5 0 6 7 6 7 6 0 5 7 6 7 8
746 1,098 0 964 756 1,173 0 1,224 756 1,117 756 1,117 0 1,224 756 1,224 756 1,224 0 1,117 756 1,173 756 1,307
3,751 4,932 – 4,530 3,805 5,267 – 5,634 3,805 5,313 3,805 5,313 – 5,634 3,805 5,634 3,805 5,634 – 5,313 3,805 5,481 3,805 5,883
68 21 – 3 6 31 – 97 66 3 4 4 – 2 2 0 10 13 – 0 0 0 156 174
20.49 (3.56–117.89) 1.72 (0.33–8.89) – (–) 0.23 (0.01–4.90) 1.03 (0.22–4.89) 1.58 (0.48–5.20) – (–) 3.82 (1.53–9.51) 1.79 (0.70–4.62) 0.06 (0.01–0.27) 0.06 (0.01–0.38) 0.03 (0.00–0.26) – (–) 0.10 (0.01–1.85) 0.13 (0.01–2.44) 0 (–) 0.22 (0.03–1.52) 0.14 (0.02–0.86) – (–) 0 (–) 0 (–) 0 (–) 8.79 (5.20–14.88) 4.70 (2.87–7.69)
P < .001 N/A .507 N/A < .001 .470 N/A .999 .286 N/A N/A < .001
*Estimated complication rate per 100 years (95% confidence interval).
Table 9 Biologic Complications with SCs Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
20 7 19 6 20 7 18 6 18 7 20 7 20 6 20 6 22 8
1,111 327 1,022 309 1,044 327 967 309 882 327 1,075 355 1,117 309 1,142 309 1,195 373
6,028 1,376 4,590 1,196 5,633 1,376 4,040 1,196 3,946 1,376 5,565 1,426 5,500 1,196 5,960 1,196 6,280 1,606
34 1 15 0 20 2 59 2 6 1 15 3 10 14 43 14 187 34
0.41 (0.19–0.88) 0.08 (0.01–0.94) 0.24 (0.08–0.71) 0 (–) 0.41 (0.18–0.93) 0.24 (0.04–1.49) 1.71 (0.53–5.57) 1.16 (0.14–9.90) 0.19 (0.05–0.69) 0.12 (0.01–1.55) 0.27 (0.16–0.47) 0.22 (0.07–0.72) 0.19 (0.06–0.58) 1.09 (0.30–4.02) 1.22 (0.41–3.65) 0.93 (0.28–3.11) 3.40 (1.98–5.82) 4.46 (2.07–9.58)
P .213 .999 .567 .687 .704 .736 .010 .698 .425
*Estimated complication rate per 100 years (95% confidence interval).
All Fixed Prostheses. Biologic complications with “all fixed prostheses” are summarized in Table 13. There were statistically significantly more overall complications with cemented prostheses (P < .001). No
individual complication reached significance, but there was a tendency toward more implant loss for cemented prostheses (P = .074).
116 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 10
Biologic Complications with FPPs
Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
2 8 2 8 2 8 1 8 1 7 2 10 2 9 2 9 2 10
25 435 25 461 25 461 16 435 16 315 25 624 25 581 25 581 25 624
150 2,572 150 2,759 150 2,759 48 2,572 48 1,372 150 4,174 150 3,959 150 3,959 150 4,174
0 28 0 10 0 11 0 59 0 0 0 33 0 9 0 32 0 149
0 (–) 1.35 (0.25–7.27) 0 (–) 0.42 (0.13–1.34) 0 (–) 0.38 (0.16–0.90) 0 (–) 2.04 (0.83–5.02) 0 (–) 0 (–) 0 (–) 0.76 (0.34–1.70) 0 (–) 0.17 (0.04–0.77) 0 (–) 0.98 (0.20–4.84) 0 (–) 4.18 (2.11–8.28)
P > .999 > .999 > .999 .99 N/A > .999 > .999 > .999 > .999
*Estimated complication rate per 100 years (95% confidence interval).
Table 11 Biologic Complications with FAFDPs Complication
No. of No. of Exposure No. of Retention studies prostheses time (y) complications
Complication rate* (95% CI)
Bone loss (> 2 mm) Peri-implantitis
Screw Screw
13 11
742 651
6,121 5,811
356 0
3.28 (1.14–9.44) 0 (–)
Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
Screw Screw Screw Screw Screw Screw Screw
11 12 11 13 11 12 14
651 680 651 729 651 713 758
5,811 5,935 5,811 6,045 5,811 5,997 6,169
23 100 0 70 1 161 641
0.62 (0.06–6.97) 1.61 (0.56–4.62) 0 (–) 2.21 (0.73–6.71) 0.03 (0.00–0.51) 2.94 (1.36–6.33) 7.61 (4.55–12.72)
*Estimated complication rate per 100 years (95% confidence interval).
Outcome of Multivariate Poisson Regression Analysis
Prosthesis Failure The results and incidence rate ratios for this analysis are presented in Table 14. There was no statistically significant difference between the failure rates for cemented and screw-retained prostheses. Technical Complications. The results are presented in Table 15. There was a statistically significantly greater incidence of technical complications in cemented compared to screw-retained prostheses (P = .01). FAFDPs (P < .001), cFDPs (P < .001), and “all fixed prostheses” (P = .003) had significantly higher complication rates than SCs. The complication rates of acrylic resin, ceramic, and “all materials” did not dif-
fer significantly from that of porcelain-fused-to-metal (PFM) prostheses. The complication rates of gold and ceramic abutment materials did not differ significantly from that of titanium abutments. Titanium abutments showed a tendency toward higher complication rates than the “all abutment materials” group (P = .051). Biologic Complications. The results are presented in Table 1. There was a statistically significantly greater incidence of biologic complications in cemented than in screw-retained prostheses (P = .001, IRR 1.73). “All fixed prostheses” had a statistically significantly higher complication rate than SCs (P = .032). Cantilever prostheses showed a tendency to have a higher complication rate than SCs (P = .051). The International Journal of Oral & Maxillofacial Implants 117
© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 12 Biologic Complications with cFDPs No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI)
Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
1 0 1 0 1 0 0 0 0 0 1 1 1 0 1 0 1 1
15 0 15 0 15 0 0 0 0 0 15 23 15 0 15 0 15 23
169.5 – 169.5 – 169.5 – – – – – 169.5 115.0 169.5 – 169.5 – 169.5 115.0
0 – 9 – 0 – – – – – 2 1 0 – 0 – 9 0
0.00 N/A 5.31 N/A 0.00 N/A N/A N/A N/A N/A 1.18 0.87 0.00 N/A 0.00 N/A 5.31 0.00
*Estimated complication rate per 100 years (95% confidence interval).
Table 13 Biologic Complications for “All Fixed Prostheses” Complication Bone loss (> 2 mm) Peri-implantitis Presence of fistula, suppuration Peri-implant mucositis Recession Loss of implant Any esthetic problem Other Summary of all complications except “loss of implant”
No. of No. of Exposure No. of Complication rate* Retention studies prostheses time (y) complications (95% CI) Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw Cement Screw
6 5 6 5 6 5 6 4 6 4 6 6 6 4 6 5 6 6
629 1,128 629 1,128 629 1,128 629 1,072 629 1,072 629 1,235 629 1,072 629 1,128 629 1,235
3,149 5,345 3,149 5,345 3,149 5,345 3,149 5,177 3,149 5,177 3,149 5,666 3,149 5,177 3,149 5,345 3,149 5,666
5 85 22 38 35 0 1 6 0 0 40 76 7 0 2 0 72 129
P
2.89 (0.52–16.09) .50 4.27 (1.11–16.39) 1.74 (0.38–7.94) .18 1.19 (0.28–5.10) 1.27 (0.04–44.33) > .999 0 (–) 0.06 (0.01–0.72) .27 0.21 (0.04–1.18) 0 (–) N/A 0 (–) 2.29 (0.90–5.79) .07 1.54 (0.65–3.63) 0.30 (0.02–4.77) > .999 0 (–) 0.10 (0.01–1.07) .99 0 (–) 12.89 (3.45–48.16) < .001 5.47 (1.53–19.59)
*Estimated complication rate per 100 years (95% confidence interval).
DISCUSSION Restoring partially edentulous patients with the use of dental implants has been reported as a successful treatment approach, with high success and survival rates for both implants and implant-supported prostheses.83,84 Many factors, including the clinician’s preference,
influence the choice of retention. It is plausible to restore a single edentulous gap with an implant-supported SC, but to rehabilitate an extended edentulous gap, the use of several implants might be considered, depending on the planned prosthetic outcome. In this situation, a fixed prosthesis may consist of solely implant-supported FPP, a cFDP, an FAFDP, or several implant-supported SCs.
118 Volume 30, Number 1, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Millen et al
Table 14 Multivariate Analysis of Failure Estimated IRR Retention Cement vs screw Type of prosthesis FPD+cantilever vs SC FAFDP vs SC All fixed prostheses vs SC Prosthesis material Acrylic vs PFM Ceramic vs PFM All materials vs PFM Abutment material Gold vs titanium Ceramic vs titanium All abutment materials vs titanium
Table 15 Multivariate Analysis of Technical Complications 95% CI
P
0.92
0.50–1.69
.794
1.91 1.63 0.41
0.86–4.27 0.49–5.39 0.14–1.17
.113 .424 .096
0.60 1.34 3.45
0.16–2.29 0.38–4.74 1.44–8.26
.457 .654 .005
2.24 1.58 0.58
0.71–7.05 0.43–5.83 0.19–1.78
.167 .491 .341
The purpose of this investigation was to identify the influence of fixed prosthesis and retention type on technical and biologic complication rates. The European Association of Osseointegration (EAO) and the International Team for Implantology (ITI) recently published consensus statements and clinical recommendations regarding the question of whether screw or cement retention is preferred for fixed prostheses.85,86 These two sets of statements and recommendations complement one another and indicate a preference for screw retention based on retrievability and a higher frequency of biologic complications with cemented prostheses. The complication rates of each prosthesis type in the current review revealed a variety of different outcomes. There was a tendency toward more technical complications with screw-retained SCs, without statistical significance. More specifically, there were significantly more instances of abutment loosening with screw-retained crowns, which was in accordance with the results of the systematic review of Sailer et al.1 More framework fractures occurred with cemented SCs; however, this could not be compared to Sailer et al, as data on framework fracture were not presented in their published article. No statistically significant differences were identified for biologic complications relative to retention type for single implant–supported crowns, whereas in the previous review from Sailer et al,1 more soft tissue complications were documented with screw-retained SCs. This difference may have resulted because of esthetic complications that, while not considered by Sailer et al, contributed to the overall biologic complications in the current review. Esthetic complications may be of particular importance to patients.87
Retention Cement vs screw Type of prosthesis FPP vs SC FAFDP vs SC cFDP vs SC All fixed prostheses vs SC Prosthesis material Acrylic vs PFM Ceramic vs PFM All materials vs PFM Abutment material Gold vs titanium Ceramic vs titanium All abutment materials vs titanium
Estimated IRR
95% CI
1.36
1.08–1.72
1.58 9.80
0.88–2.84 .127 3.64–26.38
