Lars Schropp Ann Wenzel Andreas Stavropoulos
Authors’ affiliations: Lars Schropp, Prosthetic Dentistry, Department of Dentistry, Aarhus University, Aarhus, Denmark Lars Schropp, Ann Wenzel, Oral Radiology, Department of Dentistry, Aarhus University, Aarhus, Denmark Ann Wenzel, Radiology, Department of Odontology, University of Copenhagen, Copenhagen, Denmark Andreas Stavropoulos, Department of Periodontology and Community Dentistry, Malm€ o University, Malm€ o, Sweden Corresponding author: Lars Schropp, DDS, PhD Prosthetic Dentistry & Oral Radiology, Department of Dentistry, Aarhus University, Vennelyst Boulevard 9, 8000 Aarhus C, Denmark Tel.: +0045 87 16 80 87 Fax: +0045 86 19 56 65 e-mail: [email protected]
Early, delayed, or late single implant placement: 10-year results from a randomized controlled clinical trial
Key words: delayed implantation, dental implant, early implantation, late implantation,
randomized controlled trial, single-tooth Abstract Aim: The aim of this study was to present the 10-year clinical and radiographic data from a RCT on single-tooth implants placed early, delayed, or late after tooth extraction. Materials and Methods: Sixty-three patients were randomly allocated to three groups and received an implant on average 10 days (Ea), 3 months (De), or 17 months (La) after tooth extraction. Second-stage surgery was performed after 3 months of submerged healing; metal-ceramic crowns were cemented after one additional month. Standardized periapical radiographs were taken 1 week after implant placement (TP), 1 week (TC) and 1–1.5 year (T1) after crown delivery, and 10 years after implant placement (T10). Pocket depth (PD) and bleeding on probing were registered during controls (TC – T10). Results: Two Ea and one De implants failed to osseointegrate. Seven patients (4 Ea, 1 De, and 2 La) were not available at T10. No significant differences were found among groups regarding implant survival or radiographic peri-implant marginal bone levels (Ea: 1.15 0.77; De: 1.53 1.06; La: 1.42 1.07) at T10. Similarly, no differences were observed among groups in the number of implants with PD ≥5 mm (Ea: 29%; De: 35%; La: 44%) or the average depth of the sites with PD ≥5 mm (Ea: 5.4 0.7; De: 6.1 1.4; La: 5.4 0.5) at T10. Peri-implant mucositis was found in 70% of the cases; peri-implantitis was diagnosed only in two implants (1 De, 1 La) corresponding to 4.3%. Conclusion: Single-tooth implants placed early or delayed after tooth extraction show high survival rates and limited peri-implant marginal bone resorption or biological complications, similar to what is observed with implants placed according to the conventional (late) protocol.
Date: Accepted 27 August 2013 To cite this article: Schropp L, Wenzel A, Stavropoulos A. Early, delayed, or late single implant placement: 10-year results from a randomized controlled clinical trial. Clin. Oral Impl. Res. 25, 2014, 1359–1365 doi: 10.1111/clr.12273
Dental implants have been used for replacing missing teeth during the last four decades and are nowadays considered a reliable treatment option. A drawback of the original protocol of implant therapy (Br anemark 1985), compared with, for example, fabrication of a tooth-retained fixed partial prosthesis, was the longer course of treatment. This was due to the belief that, to achieve “proper” osseointegration, complete healing of the alveolar bone after tooth extraction is needed before implant placement; thus, a post-extraction healing period of 6 months or more was suggested (Adell et al. 1981). One concern regarding placing an implant just after tooth extraction was that lack of adequate implant stability due to the presence of a gap between the alveolar bone wall and the implant— because of shape and/or size differences between the two—or the presence of the gap
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
per se would result in partial or complete fibrous encapsulation of the implant (Carlsson et al. 1988; Caudill & Meffert 1991; Knox et al. 1991). Another concern was that infection related to the extracted tooth would jeopardize osseointegration (Becker & Becker 1990; Lundgren & Nyman 1991; Werbitt & Goldberg 1992). Increased understanding of the biology of osseointegration (Br anemark 1985), however, has indicated that the above concept of complete extraction socket healing could be challenged and that implant insertion even immediately after removal of the tooth could be performed. Preclinical in vivo experiments on implant placement into extraction sockets have indeed shown adequate osseointegration despite the presence of a gap between the alveolar bone wall and the implant at installation (Araujo et al. 2005, 2006a,b) or the
1359
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
presence of infection prior to extraction (Novaes Junior & Novaes 1995; Novaes Junior et al. 1998). Similarly, in the clinic, the gap between the alveolar bone wall and the implant at installation largely disappears through bone fill, but does not completely prevent resorption of the buccal/lingual bone during the osseointegration phase (Schropp et al. 2003; Botticelli et al. 2004; Covani et al. 2004; De Rouck et al. 2008). The 3rd International Team for Implantology (ITI) Consensus Conference categorized the various alternative protocols regarding timing of implant placement relative to tooth extraction as follows: type 1 (immediate) when the implant is installed within the same session as the tooth extraction; type 2 (early) when the implant is installed within 1–2 months after tooth extraction; type 3 (delayed), when the implant is installed within 3–4 months after tooth extraction; and type 4 (late = conventional), when the implant is installed more than 4 months after extraction (H€ammerle et al. 2004). Recent narrative and systematic reviews of clinical studies involving immediate, early, or delayed implant installation in extraction sockets (Chen et al. 2004; Quirynen et al. 2007; den Hartog et al. 2008; Lang et al. 2012; Sanz et al. 2012) have shown that such alternative protocols can lead to high survival rates of the implants, similar to those obtained with the conventional (i.e., late) installation protocol (Jung et al. 2008, 2012), at least on the short term. However, from the above-mentioned systematic reviews (Lang et al. 2012; Sanz et al. 2012), it became apparent that besides the short follow-up time of the vast majority of studies, information regarding peri-implant tissue conditions and biological complications has been scarce. Clearly, information on the long-term performance, including survival rates, peri-implant tissue conditions, and biological complication rates, is needed if conclusions on the potential benefits of the various alternative timing protocols (immediate or postponed implant placement) compared with late implant placement are to be drawn. Therefore, the aim of this report was to present 10year clinical and radiographic data from a randomized controlled clinical trial of early-, delayed-, or late-placed single-tooth implants.
Material and methods Patients and randomization
Seventy-two patients referred to the Department of Dentistry, Aarhus University, Denmark, for upper or lower single-tooth extraction and implant treatment were at
1360 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
the screening examination randomly allocated to three groups for early, delayed, or late implant placement by drawing a lot from an envelope containing 24 lots of each group. Nine patients were later either judged not suitable for single implant therapy or withdrew during the period from tooth extraction to commencement of implant treatment (Fig. 1). Thus, 63 patients (35 women, 28 men) with a mean age of 46 years (range 20–74) were treated with a single implant from 1999 to 2002. Tooth extraction was due to advanced carious, endodontic, or periodontal lesions, root fracture, or combinations thereof. Patients with a history of periodontitis had been previously treated, and their disease was considered as under control before commencement of the implant therapy. Implants were placed approximately 10 days (10.8 4.1 days; early group), approximately 3 months (98.0 19.1 days; delayed group), or approximately 1.5 years (16.7 6.9 months; late group) after tooth extraction. The distribution of implant sites for the three time groups is presented in Fig. 2. The study was approved by The Central Denmark Region Committees on Health Research Ethics (#1999/4484). Surgical procedures
First, the teeth were carefully extracted to minimize the trauma to the bone walls of the alveolus and the sites were debrided. One Osseotiteâ Parallel-Walled implant (Biomet/
3i, Palm Beach Gardens, FL, USA) was then placed in each patient early, delayed, or late after removal of the tooth according to the randomization scheme. Full-thickness flaps were raised, and the implants were placed with the top of the cover screw as much as possible flush with the proximal bone crest. Intrabony defects, due to mismatch between the implant and extraction socket shape and size, were left for spontaneous healing in the early and delayed groups. In the delayed and late groups, autogenous bone chips harvested from neighboring sites were grafted to cover exposed implant threads in case of dehiscences or fenestrations, while this was not done in the early group. Primary wound closure was achieved by the mobilization of the buccal flap. The patients received amoxicillin 750 mg and naproxen 500 mg 1 h preoperative and during the following 5 days (3 or 2 times daily, respectively). The implants were placed by two experienced surgeons using the same surgical protocol. Second-stage surgery, for healing abutment (EP, Biomet 3i) placement, was performed by flap elevation approximately 3 months after implant installation in all groups. Grafting with autogenous bone chips was performed again in case of dehiscences or fenestrations— but not for intrabony defects—in all groups. Four to 6 weeks later, an impression was made and a metal-ceramic crown was fitted on an STA or UCLA abutment (Biomet 3i), tightened to the implant with a Gold-Tite square uniscrew (Biomet 3i) using a torque
72 allocated to randomization
24 early 1 treatment abandoned after radiography
1 included in a removable prosthesis
22 single implant crowns
4 dropouts before 10-yr
2 implant failures
24 delayed 1 included in a partial fixed prosthesis
1 implant abandoned at surgery
22 single implant crowns
1 dropout before 10-yr
1 implant failure
24 late 5 not interested in treatment
19 single implant crowns
2 dropouts before 10-yr
2 died 1 did not show 1 not found
1 did not show
1 sick 1 not interested to participate
16 at 10-yr control
20 at 10-yr control
17 at 10-yr control
2 telephone interview (survival data) 14 clinical examination 14 radiographic examination
2 telephone interview (survival data) 18 clinical examination 18 radiographic examination
17 clinical examination 16 radiographic examination
Fig. 1. Flow diagram of the patients enrolled in the study.
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
12 10 Early
Delayed
Late
8 6 4 2 0
Anterior
Premolar maxilla
Molar
Anterior
Premolar mandible
Molar
Fig. 2. Distribution of implant sites enrolled in the study.
driver (32 Ncm). All crowns were cemented with either a zinc phosphate cement (Dentsply; Detrey, Konstanz, Germany) or a temporary cement (TempBond; Kerr, Orange, CA, USA), except in two cases where the crowns replacing mandibular incisors were screwretained. The second-stage surgery and the prosthetic procedures were conducted by two experienced dentists.
Control visits
Standardized periapical radiographs of the implants were taken with an analog film (Kodak Insight; Kodak, Rochester, NY, USA) approximately 1 week after implant placement (TP) and 1 week after mounting of the implant crown (TC) and with a digital storage phosphor plate system (Digora, Soredex, Tuusula, Finland) approximately 1–1.5 years after crown delivery (T1) and approximately 10 years (mean 9.7 0.38 years) after implant placement (T10). The analog radiographs were subsequently digitized with a flatbed scanner (Hewlett-Packard; Palo Alto, CA, USA). Marginal bone levels (BLs) mesially and distally to the implant were recorded (monitor: Lenovoâ Thinkvisionâ 21-inch monitor (Lenovo, Morrisville, NC, USA) with a resolution of 1680 9 1050 pixels) by measuring linearly the distance from the implant shoulder to the first visible bone-to-implant contact using a dedicated computer software (Gotfredsen et al. 1999). Furthermore, the radiographic implant width was measured and compared with the true implant width for adjusting BL measurements according to image magnification. To estimate the reproducibility of the method, recordings were performed in duplicate in 18% of the images, with a 1-week interval. All radiographic measurements were carried out by one observer blinded to the implant group. A clinical examination including registration of probing pocket depths (PDs) ≥4 mm and bleeding/pus on probing (recorded as presence/absence) at six sites (mesio-, mid-,
and distobuccally; mesio-, mid-, and distopalatally/distolingually) of the implant was made at the various control visits. Information on smoking habits was collected; patients were classified as current smokers if they have smoked 100 cigarettes in their lifetime and smoked cigarettes every day or some days during the time of the 10-year control (Schoenborn & Adams 2010). In addition, information on any treatment related to biological complications associated with the implant, other than routine supportive periodontal therapy or prophylaxis, was recorded. Four examiners, all blinded to the treatment group, performed the clinical examinations at the various time points. Specifically, two examiners performed the clinical examination at the 10-year followup; these examiners were beforehand calibrated (data not shown) with the two other examiners who were involved in the previous control visits. Data analysis
Mesial and distal BL values at each implant were averaged, and descriptive statistics were calculated for the three groups. Comparisons between the three groups were performed by Kruskal–Wallis test, while changes over time were analyzed by Wilcoxon signed rank tests. Reproducibility of BL measurements was tested with the Wilcoxon signed rank tests.
Only PD values from mid-buccal, midlingual, and the largest value from the mesiobuccal/mesiolingual and distobuccal/distolingual sites were used for the analyses (i.e., four sites per implant). Percentage of implants with PD ≥5 mm and frequency distribution and average PD for the sites with PD ≥5 mm at TC, T1, and T10 were calculated for each treatment group. For comparisons among control visits, only the data from patients attending all examinations were included; differences in percentage of implants or sites with PD ≥5 mm between the three groups were tested with Pearson’s chi-square tests. The level of statistical significance was set at P < 0.05.
Results Fate of the implants and patient dropout from time of randomization to T10 are shown in Fig. 1. Three implants (2 early and 1 delayed) failed to osseointegrate in the period from implant placement to second-stage surgery (i.e., before loading) and were explanted at reentry, while no implants were lost hereafter. Seven patients (4 early, 1 delayed, and 2 late) were not available at T10 for different reasons (i.e., dropout of 11.7%). Mean age of the 53 patients (31 women, 22 men) available at T10 was 55.2 years. Information on implant survival for four patients was gained only through a telephone interview. Thus, clinical data at T10 were available from 49 patients, while 47 patients attended all control visits. Radiographic data were available from 46 patients at TP, 47 at TC, 41 at T1, and from 48 patients at T10. Six patients in the late group missed their radiographic examination at T1 due to long time closure because of rebuilding of the Section of Oral Radiology at the Department of Dentistry during that time. The distribution of implant sites in the three treatment groups for patients attending the 10-year control visit is presented in Fig. 3. No patients reported
8 7 Early
6
Delayed
Late
5 4 3 2 1 0
Anterior
Premolar maxilla
Molar
Anterior
Premolar mandible
Molar
Fig. 3. Distribution of implant sites for patients attending the 10-year control visit.
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
1361 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
and late group (1.09 and 0.99 mm, respectively). Table 2 presents BL changes over time. For all implants, a significant bone loss of on average 0.66 mm was seen from implant placement to TC. Bone loss was least (0.3 mm) in the early group, while in the delayed and late groups, BL at TC was located significantly more apically compared with TP (0.5 mm and 1.2 mm, respectively). During the period from TP to T1, for all implants, a significant bone loss of 0.54 mm was seen. It should be noted that radiographs from only eight patients in the late group were available for this analysis. No significant differences in BL were found between TC and the 10-year control visit (T10) or between T1 and T10, on average for all implants, but also for each group.
having received any other treatment than standard supportive periodontal therapy or prophylaxis. Five of 14 patients in the early group, 8 of 17 in the delayed group, and 3 of 16 in the late group were classified as smokers. Radiographic evaluation
A good agreement between the first and second recordings of the marginal bone levels mesially and distally (P = 0.38/P = 0.27) and the implant width (P = 0.83) was found. The difference between the first and second measurements averaged 0.40 0.46 mm. The peri-implant marginal bone levels at implants placed early, delayed, or late after tooth extraction, at the various control visits, are presented in Table 1 and Fig. 4. No significant differences between the three groups were found 10 years after implant placement (P = 0.56), and BL for all 48 implants was on average located 1.38 mm apically to the implant shoulder. Differences between the three groups were only found at T1 (P = 0.004), where BL was located more apically at the implants in the delayed group (1.71 mm) compared with those in the early
Clinical evaluation
Table 3 shows the percentage of implants with at least one site with PD ≥5 mm and the site (buccal, palatal/lingual, and mesial and distal) frequency distribution and mean/ median PD for the sites with PD ≥5 mm, in the three groups. At TC, 26 of the 47 patients
Table 1. Peri-implant radiographic proximal bone level (averaged for mesial and distal sites) at implants placed early (Ea), delayed (De), or late (De), at the various control visits Time
Group
N
Mean SD (mm)
TP
Ea De La All Ea De La All Ea De La All Ea De La All
14 18 14 46 14 17 16 47 14 17 10 41 14 18 16 48
0.81 0.86 0.40 0.71 1.15 1.39 1.59 1.39 1.09 1.71 0.99 1.32 1.15 1.53 1.42 1.38
TC
T1
T10
1.05 0.89 0.55 0.87 0.82 0.44 0.65 0.65 0.44 0.58 0.61 0.63 0.77 1.06 1.07 0.98
Median (mm)
P
0.33 0.54 0.18 0.33 1.27 1.42 1.57 1.43 1.12 1.64 1.08 1.31 1.00 1.54 1.54 1.27
0.36
0.24
0.004
0.56
TP, 1 week after implant placement; TC, 1 week after crown delivery; T1, 1–1.5 years after crown delivery; T10, 10 years after implant placement.
1.8 1.6 1.4
All Early Delayed Late
1.2 1 0.8 0.6 0.4 0.2 0 TP
TC
T1
Fig. 4. Peri-implant bone level (mean in mm) at TP, TC, T1, and T10.
1362 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
T10
(55%) had at least one implant site with a PD ≥5 mm; the three groups differed significantly (P < 0.01), with more sites with PD ≥5 mm in the early group and fewer in the late group (55% and 6%, respectively). One year after crown delivery (T1), 22 patients (47%) showed a PD ≥5 mm at one or more implant sites; the late group differed significantly from the early and the delayed groups (P < 0.005), the former having approximately three and five times, respectively, fewer sites with PD ≥5 mm. At the 10-year control (T10), 70% of all implants showed BoP; in particular, 64%, 94%, and 53% of the implants in the early, delayed, and late groups, respectively, exhibited BoP. Seventeen of the implants (36%) showed one or more sites with PD ≥5 mm, and there were no significant differences between the early, delayed, and late groups regarding the number of sites with a PD ≥5 mm (range 14–21%) (P = 0.65). The percentage of sites with PD ≥5 mm in the early group was reduced gradually from TC (55%) to T10 (14%), while almost no difference between the various time points was seen for the delayed group. In the late group, the percentage of sites with PD ≥5 mm was almost tripled from TC (6%) and T1 (8%) to 10-year follow-up (20%). At 10 years (T10), two implants (1 delayed, 1 late) (4.3%) suffered from peri-implantitis, defined as PD ≥5 mm, bleeding/or pus on probing, and marginal bone loss of >1 mm compared with BL at baseline, irrespective of whether TC or T1 was chosen as baseline herein.
Discussion The results of the present clinical study demonstrated that there were no differences in the long-term performance of single-tooth implants placed early, delayed, or late group after tooth extraction. High implant survival rates were observed irrespective of the time between tooth extraction and implant placement 10 years after surgery (early: 91%, delayed: 95%, and late: 100%). These implant survival rates compare well with those reported in previously published narrative and systematic reviews about the influence of timing of implant placement after tooth extraction on the outcome of implant therapy (Chen et al. 2004; Quirynen et al. 2007; den Hartog et al. 2008; Sanz et al. 2012; De Bruyn et al. 2013). In a recent systematic review on late-placed single-tooth implants, the 5- and 10-year implant survival rate was 97.2% and 95.2%, respectively (Jung
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
Table 2. Changes (in mm) in peri-implant radiographic proximal bone level (averaged for mesial and distal sites) from TP to TC, from TC to T10, as well as from TP to T1 and from T1 to T10 at implants placed early (Ea), delayed (De), and late (La) N Group
TP–TC
Ea De La All
14 17 14 45
Mean SD
Median
P
Mean SD
P
Median
TC–T10 0.34 0.49 1.20 0.66
1.52 0.90 0.84 1.15
0.91 0.87 1.28 1.02
0.30 0.04 0.00 0.00
TP–T1 Ea De La All
N
14 17 16 47
0.00 0.23 0.17 0.02
0.92 0.97 0.77 0.89
0.17 0.21 0.23 0.15
0.78 0.49 0.35 0.72
0.06 0.09 0.28 0.05
0.51 0.90 0.73 0.74
0.06 0.09 0.13 0.00
0.88 0.33 0.17 0.92
T1–T10
14 17 8 39
0.27 0.80 0.46 0.54
1.17 0.93 0.79 1.00
0.69 1.17 0.32 0.76
0.14 0.00 0.16 0.00
14 17 10 41
TP, 1 week after implant placement; TC, 1 week after crown delivery; T1, 1–1.5 years after crown delivery; T10, 10 years after implant placement.
Table 3. Percentage of implants with at least one site with PD ≥5 mm, and frequency distribution and mean SD and median PD (in mm) for the sites with PD ≥5 mm, only for those patients attending all control visits Group Time Implants TC T1 T10 Sites TC
T1
T10
All implants (N = 47)
Early (N = 14)
Delayed (N = 17)
55% (N = 26) 47% (N = 22) 36% (N = 17)
86% (N = 12) 57% (N = 8) 29% (N = 4)
59% (N = 10) 65% (N = 11) 35% (N = 6)
All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median
55% 36% 50% 68% 5.7 1.2 mm 5.0 27% 14% 7% 43% 5.6 0.8 mm 5.0 14% 7% 7% 21% 5.4 0.7 mm 5.0
32% 29% 18% 41% 5.3 0.6 mm 5.0 43% 29% 29% 56% 5.2 0.5 mm 5.0 21% 6% 6% 35% 6.1 1.4 mm 6.0
et al. 2012). In the most recent systematic review on immediate implants, an estimated 2–4-year implant survival rate of 98.4 and 97.5%, respectively, was reported (Lang et al. 2012), while in the systematic review on early or delayed implant placement (Sanz et al. 2012), an implant survival rate of 95–97.5% up to 5 years was reported. Nevertheless, it must be pointed out that 5 of the 8 included publications in the latter review (Sanz et al. 2012) were partly based on the patient material of the current report, while the three remaining studies (Nemcovsky et al. 2000; Nemcovsky & Artzi 2002; Cosyn & De Rouck 2009) had basically a relatively short follow-up period (1 year (average 2.5 years), the radiographic peri-implant marginal bone level was on average 1.3 mm apically to the implant–abutment junction (Eghbali et al. 2012). These values are close to those reported in several long-term studies for the majority of implants with similar design to that used herein (Jemt & Johansson 2006; Roccuzzo et al. 2010; Zetterqvist et al. 2010). In this context, it has to be pointed out that when evaluating peri-implant marginal bone levels among studies, both the implant system (e.g., implant design, prosthetic connection) and the time from implant installation appear to be important parameters. A recent metaanalysis on prospective studies reporting on
1363 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
three major implant systems after at least 5 years in function indicated that marginal bone loss due to remodeling indeed seems dependent, among other factors, also upon implant system design characteristics. The major part of the marginal bone loss occurred early and up to the first year post-loading, while the bone level more or less stabilized thereafter (Laurell & Lundgren 2011). In this meta-analysis, a weighted mean marginal bone level change of 0.7 mm over 5 years was calculated for implants with a design basically identical to that in the present study, challenging thus the “classic” success criteria for implant therapy, accepting bone loss not exceeding 1.5 mm within the first year after insertion of the restoration and 0.2 mm bone loss annually hereafter (Albrektsson et al. 1986; Albrektsson & Isidor 1994). The amount of peri-implant marginal bone loss found in the recent meta-analysis of Laurell and Lundgren (Laurell & Lundgren 2011) compares well with that observed in the present study. Bone loss in the three groups ranged from 0.3 to 1.0 mm, from implant insertion to the 10-year control; the largest amount of loss indeed occurred up to 1 year post-loading, and only negligible changes were observed thereafter. Mere reporting of radiographic peri-implant marginal bone level values or bone loss after some years of function even if coupled with clinical data on PD and detectable inflammation is not necessarily informative regarding the presence/absence of peri-implantitis, unless these values are compared with marginal bone level values from a relevant baseline. Herein, both the time of crown delivery (TC) and the 1-year post-loading (T1) were used as baseline in two separate comparisons (Table 2), and the threshold for BL indicative of peri-implantitis was set to >1 mm, corresponding to 2–3 times the SD of the measurement error of the radiographic analysis observer (0.46 mm)—as suggested in the
consensus report of the 8th European Workshop on Periodontology (Sanz & Chapple 2012). In the present study, only the same two implants presented with the radiographic and clinical characteristics defined above, irrespective of which time point was used as baseline, corresponding to 4% peri-implantitis. Considering the high peri-implant mucositis rate (70%) and the fact that periimplantitis rate is based on the bone loss evaluation only interproximally, it may be argued that peri-implantitis in the present study might have been underestimated and that more cases with peri-implantitis would have been diagnosed if evaluation of the buccal and palatal/lingual BL changes was also included. However, the peri-implantitis rate observed is similar to the 5-year peri-implantitis rate for cemented single-crown reconstructions (2.8%) calculated in the recent meta-analysis (Jung et al. 2012), as well to the peri-implantitis rate (
Authors’ affiliations: Lars Schropp, Prosthetic Dentistry, Department of Dentistry, Aarhus University, Aarhus, Denmark Lars Schropp, Ann Wenzel, Oral Radiology, Department of Dentistry, Aarhus University, Aarhus, Denmark Ann Wenzel, Radiology, Department of Odontology, University of Copenhagen, Copenhagen, Denmark Andreas Stavropoulos, Department of Periodontology and Community Dentistry, Malm€ o University, Malm€ o, Sweden Corresponding author: Lars Schropp, DDS, PhD Prosthetic Dentistry & Oral Radiology, Department of Dentistry, Aarhus University, Vennelyst Boulevard 9, 8000 Aarhus C, Denmark Tel.: +0045 87 16 80 87 Fax: +0045 86 19 56 65 e-mail: [email protected]
Early, delayed, or late single implant placement: 10-year results from a randomized controlled clinical trial
Key words: delayed implantation, dental implant, early implantation, late implantation,
randomized controlled trial, single-tooth Abstract Aim: The aim of this study was to present the 10-year clinical and radiographic data from a RCT on single-tooth implants placed early, delayed, or late after tooth extraction. Materials and Methods: Sixty-three patients were randomly allocated to three groups and received an implant on average 10 days (Ea), 3 months (De), or 17 months (La) after tooth extraction. Second-stage surgery was performed after 3 months of submerged healing; metal-ceramic crowns were cemented after one additional month. Standardized periapical radiographs were taken 1 week after implant placement (TP), 1 week (TC) and 1–1.5 year (T1) after crown delivery, and 10 years after implant placement (T10). Pocket depth (PD) and bleeding on probing were registered during controls (TC – T10). Results: Two Ea and one De implants failed to osseointegrate. Seven patients (4 Ea, 1 De, and 2 La) were not available at T10. No significant differences were found among groups regarding implant survival or radiographic peri-implant marginal bone levels (Ea: 1.15 0.77; De: 1.53 1.06; La: 1.42 1.07) at T10. Similarly, no differences were observed among groups in the number of implants with PD ≥5 mm (Ea: 29%; De: 35%; La: 44%) or the average depth of the sites with PD ≥5 mm (Ea: 5.4 0.7; De: 6.1 1.4; La: 5.4 0.5) at T10. Peri-implant mucositis was found in 70% of the cases; peri-implantitis was diagnosed only in two implants (1 De, 1 La) corresponding to 4.3%. Conclusion: Single-tooth implants placed early or delayed after tooth extraction show high survival rates and limited peri-implant marginal bone resorption or biological complications, similar to what is observed with implants placed according to the conventional (late) protocol.
Date: Accepted 27 August 2013 To cite this article: Schropp L, Wenzel A, Stavropoulos A. Early, delayed, or late single implant placement: 10-year results from a randomized controlled clinical trial. Clin. Oral Impl. Res. 25, 2014, 1359–1365 doi: 10.1111/clr.12273
Dental implants have been used for replacing missing teeth during the last four decades and are nowadays considered a reliable treatment option. A drawback of the original protocol of implant therapy (Br anemark 1985), compared with, for example, fabrication of a tooth-retained fixed partial prosthesis, was the longer course of treatment. This was due to the belief that, to achieve “proper” osseointegration, complete healing of the alveolar bone after tooth extraction is needed before implant placement; thus, a post-extraction healing period of 6 months or more was suggested (Adell et al. 1981). One concern regarding placing an implant just after tooth extraction was that lack of adequate implant stability due to the presence of a gap between the alveolar bone wall and the implant— because of shape and/or size differences between the two—or the presence of the gap
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
per se would result in partial or complete fibrous encapsulation of the implant (Carlsson et al. 1988; Caudill & Meffert 1991; Knox et al. 1991). Another concern was that infection related to the extracted tooth would jeopardize osseointegration (Becker & Becker 1990; Lundgren & Nyman 1991; Werbitt & Goldberg 1992). Increased understanding of the biology of osseointegration (Br anemark 1985), however, has indicated that the above concept of complete extraction socket healing could be challenged and that implant insertion even immediately after removal of the tooth could be performed. Preclinical in vivo experiments on implant placement into extraction sockets have indeed shown adequate osseointegration despite the presence of a gap between the alveolar bone wall and the implant at installation (Araujo et al. 2005, 2006a,b) or the
1359
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
presence of infection prior to extraction (Novaes Junior & Novaes 1995; Novaes Junior et al. 1998). Similarly, in the clinic, the gap between the alveolar bone wall and the implant at installation largely disappears through bone fill, but does not completely prevent resorption of the buccal/lingual bone during the osseointegration phase (Schropp et al. 2003; Botticelli et al. 2004; Covani et al. 2004; De Rouck et al. 2008). The 3rd International Team for Implantology (ITI) Consensus Conference categorized the various alternative protocols regarding timing of implant placement relative to tooth extraction as follows: type 1 (immediate) when the implant is installed within the same session as the tooth extraction; type 2 (early) when the implant is installed within 1–2 months after tooth extraction; type 3 (delayed), when the implant is installed within 3–4 months after tooth extraction; and type 4 (late = conventional), when the implant is installed more than 4 months after extraction (H€ammerle et al. 2004). Recent narrative and systematic reviews of clinical studies involving immediate, early, or delayed implant installation in extraction sockets (Chen et al. 2004; Quirynen et al. 2007; den Hartog et al. 2008; Lang et al. 2012; Sanz et al. 2012) have shown that such alternative protocols can lead to high survival rates of the implants, similar to those obtained with the conventional (i.e., late) installation protocol (Jung et al. 2008, 2012), at least on the short term. However, from the above-mentioned systematic reviews (Lang et al. 2012; Sanz et al. 2012), it became apparent that besides the short follow-up time of the vast majority of studies, information regarding peri-implant tissue conditions and biological complications has been scarce. Clearly, information on the long-term performance, including survival rates, peri-implant tissue conditions, and biological complication rates, is needed if conclusions on the potential benefits of the various alternative timing protocols (immediate or postponed implant placement) compared with late implant placement are to be drawn. Therefore, the aim of this report was to present 10year clinical and radiographic data from a randomized controlled clinical trial of early-, delayed-, or late-placed single-tooth implants.
Material and methods Patients and randomization
Seventy-two patients referred to the Department of Dentistry, Aarhus University, Denmark, for upper or lower single-tooth extraction and implant treatment were at
1360 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
the screening examination randomly allocated to three groups for early, delayed, or late implant placement by drawing a lot from an envelope containing 24 lots of each group. Nine patients were later either judged not suitable for single implant therapy or withdrew during the period from tooth extraction to commencement of implant treatment (Fig. 1). Thus, 63 patients (35 women, 28 men) with a mean age of 46 years (range 20–74) were treated with a single implant from 1999 to 2002. Tooth extraction was due to advanced carious, endodontic, or periodontal lesions, root fracture, or combinations thereof. Patients with a history of periodontitis had been previously treated, and their disease was considered as under control before commencement of the implant therapy. Implants were placed approximately 10 days (10.8 4.1 days; early group), approximately 3 months (98.0 19.1 days; delayed group), or approximately 1.5 years (16.7 6.9 months; late group) after tooth extraction. The distribution of implant sites for the three time groups is presented in Fig. 2. The study was approved by The Central Denmark Region Committees on Health Research Ethics (#1999/4484). Surgical procedures
First, the teeth were carefully extracted to minimize the trauma to the bone walls of the alveolus and the sites were debrided. One Osseotiteâ Parallel-Walled implant (Biomet/
3i, Palm Beach Gardens, FL, USA) was then placed in each patient early, delayed, or late after removal of the tooth according to the randomization scheme. Full-thickness flaps were raised, and the implants were placed with the top of the cover screw as much as possible flush with the proximal bone crest. Intrabony defects, due to mismatch between the implant and extraction socket shape and size, were left for spontaneous healing in the early and delayed groups. In the delayed and late groups, autogenous bone chips harvested from neighboring sites were grafted to cover exposed implant threads in case of dehiscences or fenestrations, while this was not done in the early group. Primary wound closure was achieved by the mobilization of the buccal flap. The patients received amoxicillin 750 mg and naproxen 500 mg 1 h preoperative and during the following 5 days (3 or 2 times daily, respectively). The implants were placed by two experienced surgeons using the same surgical protocol. Second-stage surgery, for healing abutment (EP, Biomet 3i) placement, was performed by flap elevation approximately 3 months after implant installation in all groups. Grafting with autogenous bone chips was performed again in case of dehiscences or fenestrations— but not for intrabony defects—in all groups. Four to 6 weeks later, an impression was made and a metal-ceramic crown was fitted on an STA or UCLA abutment (Biomet 3i), tightened to the implant with a Gold-Tite square uniscrew (Biomet 3i) using a torque
72 allocated to randomization
24 early 1 treatment abandoned after radiography
1 included in a removable prosthesis
22 single implant crowns
4 dropouts before 10-yr
2 implant failures
24 delayed 1 included in a partial fixed prosthesis
1 implant abandoned at surgery
22 single implant crowns
1 dropout before 10-yr
1 implant failure
24 late 5 not interested in treatment
19 single implant crowns
2 dropouts before 10-yr
2 died 1 did not show 1 not found
1 did not show
1 sick 1 not interested to participate
16 at 10-yr control
20 at 10-yr control
17 at 10-yr control
2 telephone interview (survival data) 14 clinical examination 14 radiographic examination
2 telephone interview (survival data) 18 clinical examination 18 radiographic examination
17 clinical examination 16 radiographic examination
Fig. 1. Flow diagram of the patients enrolled in the study.
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
12 10 Early
Delayed
Late
8 6 4 2 0
Anterior
Premolar maxilla
Molar
Anterior
Premolar mandible
Molar
Fig. 2. Distribution of implant sites enrolled in the study.
driver (32 Ncm). All crowns were cemented with either a zinc phosphate cement (Dentsply; Detrey, Konstanz, Germany) or a temporary cement (TempBond; Kerr, Orange, CA, USA), except in two cases where the crowns replacing mandibular incisors were screwretained. The second-stage surgery and the prosthetic procedures were conducted by two experienced dentists.
Control visits
Standardized periapical radiographs of the implants were taken with an analog film (Kodak Insight; Kodak, Rochester, NY, USA) approximately 1 week after implant placement (TP) and 1 week after mounting of the implant crown (TC) and with a digital storage phosphor plate system (Digora, Soredex, Tuusula, Finland) approximately 1–1.5 years after crown delivery (T1) and approximately 10 years (mean 9.7 0.38 years) after implant placement (T10). The analog radiographs were subsequently digitized with a flatbed scanner (Hewlett-Packard; Palo Alto, CA, USA). Marginal bone levels (BLs) mesially and distally to the implant were recorded (monitor: Lenovoâ Thinkvisionâ 21-inch monitor (Lenovo, Morrisville, NC, USA) with a resolution of 1680 9 1050 pixels) by measuring linearly the distance from the implant shoulder to the first visible bone-to-implant contact using a dedicated computer software (Gotfredsen et al. 1999). Furthermore, the radiographic implant width was measured and compared with the true implant width for adjusting BL measurements according to image magnification. To estimate the reproducibility of the method, recordings were performed in duplicate in 18% of the images, with a 1-week interval. All radiographic measurements were carried out by one observer blinded to the implant group. A clinical examination including registration of probing pocket depths (PDs) ≥4 mm and bleeding/pus on probing (recorded as presence/absence) at six sites (mesio-, mid-,
and distobuccally; mesio-, mid-, and distopalatally/distolingually) of the implant was made at the various control visits. Information on smoking habits was collected; patients were classified as current smokers if they have smoked 100 cigarettes in their lifetime and smoked cigarettes every day or some days during the time of the 10-year control (Schoenborn & Adams 2010). In addition, information on any treatment related to biological complications associated with the implant, other than routine supportive periodontal therapy or prophylaxis, was recorded. Four examiners, all blinded to the treatment group, performed the clinical examinations at the various time points. Specifically, two examiners performed the clinical examination at the 10-year followup; these examiners were beforehand calibrated (data not shown) with the two other examiners who were involved in the previous control visits. Data analysis
Mesial and distal BL values at each implant were averaged, and descriptive statistics were calculated for the three groups. Comparisons between the three groups were performed by Kruskal–Wallis test, while changes over time were analyzed by Wilcoxon signed rank tests. Reproducibility of BL measurements was tested with the Wilcoxon signed rank tests.
Only PD values from mid-buccal, midlingual, and the largest value from the mesiobuccal/mesiolingual and distobuccal/distolingual sites were used for the analyses (i.e., four sites per implant). Percentage of implants with PD ≥5 mm and frequency distribution and average PD for the sites with PD ≥5 mm at TC, T1, and T10 were calculated for each treatment group. For comparisons among control visits, only the data from patients attending all examinations were included; differences in percentage of implants or sites with PD ≥5 mm between the three groups were tested with Pearson’s chi-square tests. The level of statistical significance was set at P < 0.05.
Results Fate of the implants and patient dropout from time of randomization to T10 are shown in Fig. 1. Three implants (2 early and 1 delayed) failed to osseointegrate in the period from implant placement to second-stage surgery (i.e., before loading) and were explanted at reentry, while no implants were lost hereafter. Seven patients (4 early, 1 delayed, and 2 late) were not available at T10 for different reasons (i.e., dropout of 11.7%). Mean age of the 53 patients (31 women, 22 men) available at T10 was 55.2 years. Information on implant survival for four patients was gained only through a telephone interview. Thus, clinical data at T10 were available from 49 patients, while 47 patients attended all control visits. Radiographic data were available from 46 patients at TP, 47 at TC, 41 at T1, and from 48 patients at T10. Six patients in the late group missed their radiographic examination at T1 due to long time closure because of rebuilding of the Section of Oral Radiology at the Department of Dentistry during that time. The distribution of implant sites in the three treatment groups for patients attending the 10-year control visit is presented in Fig. 3. No patients reported
8 7 Early
6
Delayed
Late
5 4 3 2 1 0
Anterior
Premolar maxilla
Molar
Anterior
Premolar mandible
Molar
Fig. 3. Distribution of implant sites for patients attending the 10-year control visit.
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
1361 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
and late group (1.09 and 0.99 mm, respectively). Table 2 presents BL changes over time. For all implants, a significant bone loss of on average 0.66 mm was seen from implant placement to TC. Bone loss was least (0.3 mm) in the early group, while in the delayed and late groups, BL at TC was located significantly more apically compared with TP (0.5 mm and 1.2 mm, respectively). During the period from TP to T1, for all implants, a significant bone loss of 0.54 mm was seen. It should be noted that radiographs from only eight patients in the late group were available for this analysis. No significant differences in BL were found between TC and the 10-year control visit (T10) or between T1 and T10, on average for all implants, but also for each group.
having received any other treatment than standard supportive periodontal therapy or prophylaxis. Five of 14 patients in the early group, 8 of 17 in the delayed group, and 3 of 16 in the late group were classified as smokers. Radiographic evaluation
A good agreement between the first and second recordings of the marginal bone levels mesially and distally (P = 0.38/P = 0.27) and the implant width (P = 0.83) was found. The difference between the first and second measurements averaged 0.40 0.46 mm. The peri-implant marginal bone levels at implants placed early, delayed, or late after tooth extraction, at the various control visits, are presented in Table 1 and Fig. 4. No significant differences between the three groups were found 10 years after implant placement (P = 0.56), and BL for all 48 implants was on average located 1.38 mm apically to the implant shoulder. Differences between the three groups were only found at T1 (P = 0.004), where BL was located more apically at the implants in the delayed group (1.71 mm) compared with those in the early
Clinical evaluation
Table 3 shows the percentage of implants with at least one site with PD ≥5 mm and the site (buccal, palatal/lingual, and mesial and distal) frequency distribution and mean/ median PD for the sites with PD ≥5 mm, in the three groups. At TC, 26 of the 47 patients
Table 1. Peri-implant radiographic proximal bone level (averaged for mesial and distal sites) at implants placed early (Ea), delayed (De), or late (De), at the various control visits Time
Group
N
Mean SD (mm)
TP
Ea De La All Ea De La All Ea De La All Ea De La All
14 18 14 46 14 17 16 47 14 17 10 41 14 18 16 48
0.81 0.86 0.40 0.71 1.15 1.39 1.59 1.39 1.09 1.71 0.99 1.32 1.15 1.53 1.42 1.38
TC
T1
T10
1.05 0.89 0.55 0.87 0.82 0.44 0.65 0.65 0.44 0.58 0.61 0.63 0.77 1.06 1.07 0.98
Median (mm)
P
0.33 0.54 0.18 0.33 1.27 1.42 1.57 1.43 1.12 1.64 1.08 1.31 1.00 1.54 1.54 1.27
0.36
0.24
0.004
0.56
TP, 1 week after implant placement; TC, 1 week after crown delivery; T1, 1–1.5 years after crown delivery; T10, 10 years after implant placement.
1.8 1.6 1.4
All Early Delayed Late
1.2 1 0.8 0.6 0.4 0.2 0 TP
TC
T1
Fig. 4. Peri-implant bone level (mean in mm) at TP, TC, T1, and T10.
1362 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
T10
(55%) had at least one implant site with a PD ≥5 mm; the three groups differed significantly (P < 0.01), with more sites with PD ≥5 mm in the early group and fewer in the late group (55% and 6%, respectively). One year after crown delivery (T1), 22 patients (47%) showed a PD ≥5 mm at one or more implant sites; the late group differed significantly from the early and the delayed groups (P < 0.005), the former having approximately three and five times, respectively, fewer sites with PD ≥5 mm. At the 10-year control (T10), 70% of all implants showed BoP; in particular, 64%, 94%, and 53% of the implants in the early, delayed, and late groups, respectively, exhibited BoP. Seventeen of the implants (36%) showed one or more sites with PD ≥5 mm, and there were no significant differences between the early, delayed, and late groups regarding the number of sites with a PD ≥5 mm (range 14–21%) (P = 0.65). The percentage of sites with PD ≥5 mm in the early group was reduced gradually from TC (55%) to T10 (14%), while almost no difference between the various time points was seen for the delayed group. In the late group, the percentage of sites with PD ≥5 mm was almost tripled from TC (6%) and T1 (8%) to 10-year follow-up (20%). At 10 years (T10), two implants (1 delayed, 1 late) (4.3%) suffered from peri-implantitis, defined as PD ≥5 mm, bleeding/or pus on probing, and marginal bone loss of >1 mm compared with BL at baseline, irrespective of whether TC or T1 was chosen as baseline herein.
Discussion The results of the present clinical study demonstrated that there were no differences in the long-term performance of single-tooth implants placed early, delayed, or late group after tooth extraction. High implant survival rates were observed irrespective of the time between tooth extraction and implant placement 10 years after surgery (early: 91%, delayed: 95%, and late: 100%). These implant survival rates compare well with those reported in previously published narrative and systematic reviews about the influence of timing of implant placement after tooth extraction on the outcome of implant therapy (Chen et al. 2004; Quirynen et al. 2007; den Hartog et al. 2008; Sanz et al. 2012; De Bruyn et al. 2013). In a recent systematic review on late-placed single-tooth implants, the 5- and 10-year implant survival rate was 97.2% and 95.2%, respectively (Jung
© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
Table 2. Changes (in mm) in peri-implant radiographic proximal bone level (averaged for mesial and distal sites) from TP to TC, from TC to T10, as well as from TP to T1 and from T1 to T10 at implants placed early (Ea), delayed (De), and late (La) N Group
TP–TC
Ea De La All
14 17 14 45
Mean SD
Median
P
Mean SD
P
Median
TC–T10 0.34 0.49 1.20 0.66
1.52 0.90 0.84 1.15
0.91 0.87 1.28 1.02
0.30 0.04 0.00 0.00
TP–T1 Ea De La All
N
14 17 16 47
0.00 0.23 0.17 0.02
0.92 0.97 0.77 0.89
0.17 0.21 0.23 0.15
0.78 0.49 0.35 0.72
0.06 0.09 0.28 0.05
0.51 0.90 0.73 0.74
0.06 0.09 0.13 0.00
0.88 0.33 0.17 0.92
T1–T10
14 17 8 39
0.27 0.80 0.46 0.54
1.17 0.93 0.79 1.00
0.69 1.17 0.32 0.76
0.14 0.00 0.16 0.00
14 17 10 41
TP, 1 week after implant placement; TC, 1 week after crown delivery; T1, 1–1.5 years after crown delivery; T10, 10 years after implant placement.
Table 3. Percentage of implants with at least one site with PD ≥5 mm, and frequency distribution and mean SD and median PD (in mm) for the sites with PD ≥5 mm, only for those patients attending all control visits Group Time Implants TC T1 T10 Sites TC
T1
T10
All implants (N = 47)
Early (N = 14)
Delayed (N = 17)
55% (N = 26) 47% (N = 22) 36% (N = 17)
86% (N = 12) 57% (N = 8) 29% (N = 4)
59% (N = 10) 65% (N = 11) 35% (N = 6)
All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median All sites Buccal Palatal/lingual Mesial & distal Mean PD SD Median
55% 36% 50% 68% 5.7 1.2 mm 5.0 27% 14% 7% 43% 5.6 0.8 mm 5.0 14% 7% 7% 21% 5.4 0.7 mm 5.0
32% 29% 18% 41% 5.3 0.6 mm 5.0 43% 29% 29% 56% 5.2 0.5 mm 5.0 21% 6% 6% 35% 6.1 1.4 mm 6.0
et al. 2012). In the most recent systematic review on immediate implants, an estimated 2–4-year implant survival rate of 98.4 and 97.5%, respectively, was reported (Lang et al. 2012), while in the systematic review on early or delayed implant placement (Sanz et al. 2012), an implant survival rate of 95–97.5% up to 5 years was reported. Nevertheless, it must be pointed out that 5 of the 8 included publications in the latter review (Sanz et al. 2012) were partly based on the patient material of the current report, while the three remaining studies (Nemcovsky et al. 2000; Nemcovsky & Artzi 2002; Cosyn & De Rouck 2009) had basically a relatively short follow-up period (1 year (average 2.5 years), the radiographic peri-implant marginal bone level was on average 1.3 mm apically to the implant–abutment junction (Eghbali et al. 2012). These values are close to those reported in several long-term studies for the majority of implants with similar design to that used herein (Jemt & Johansson 2006; Roccuzzo et al. 2010; Zetterqvist et al. 2010). In this context, it has to be pointed out that when evaluating peri-implant marginal bone levels among studies, both the implant system (e.g., implant design, prosthetic connection) and the time from implant installation appear to be important parameters. A recent metaanalysis on prospective studies reporting on
1363 |
Clin. Oral Impl. Res. 25, 2014 / 1359–1365
Schropp et al Early, delayed, or late single implant placement: a 10-year RCT
three major implant systems after at least 5 years in function indicated that marginal bone loss due to remodeling indeed seems dependent, among other factors, also upon implant system design characteristics. The major part of the marginal bone loss occurred early and up to the first year post-loading, while the bone level more or less stabilized thereafter (Laurell & Lundgren 2011). In this meta-analysis, a weighted mean marginal bone level change of 0.7 mm over 5 years was calculated for implants with a design basically identical to that in the present study, challenging thus the “classic” success criteria for implant therapy, accepting bone loss not exceeding 1.5 mm within the first year after insertion of the restoration and 0.2 mm bone loss annually hereafter (Albrektsson et al. 1986; Albrektsson & Isidor 1994). The amount of peri-implant marginal bone loss found in the recent meta-analysis of Laurell and Lundgren (Laurell & Lundgren 2011) compares well with that observed in the present study. Bone loss in the three groups ranged from 0.3 to 1.0 mm, from implant insertion to the 10-year control; the largest amount of loss indeed occurred up to 1 year post-loading, and only negligible changes were observed thereafter. Mere reporting of radiographic peri-implant marginal bone level values or bone loss after some years of function even if coupled with clinical data on PD and detectable inflammation is not necessarily informative regarding the presence/absence of peri-implantitis, unless these values are compared with marginal bone level values from a relevant baseline. Herein, both the time of crown delivery (TC) and the 1-year post-loading (T1) were used as baseline in two separate comparisons (Table 2), and the threshold for BL indicative of peri-implantitis was set to >1 mm, corresponding to 2–3 times the SD of the measurement error of the radiographic analysis observer (0.46 mm)—as suggested in the
consensus report of the 8th European Workshop on Periodontology (Sanz & Chapple 2012). In the present study, only the same two implants presented with the radiographic and clinical characteristics defined above, irrespective of which time point was used as baseline, corresponding to 4% peri-implantitis. Considering the high peri-implant mucositis rate (70%) and the fact that periimplantitis rate is based on the bone loss evaluation only interproximally, it may be argued that peri-implantitis in the present study might have been underestimated and that more cases with peri-implantitis would have been diagnosed if evaluation of the buccal and palatal/lingual BL changes was also included. However, the peri-implantitis rate observed is similar to the 5-year peri-implantitis rate for cemented single-crown reconstructions (2.8%) calculated in the recent meta-analysis (Jung et al. 2012), as well to the peri-implantitis rate (
