Statement of problem
Despite an overall high survival rate for dental implants, the effectiveness of implant retreatment remains unclear.
The purpose of this systematic review was to examine the survival rate of implants placed at sites which had an implant failure and to investigate factors that might affect outcomes after retreatment.
Material and methods
A search of electronic databases limited to English language articles was conducted using the following MeSH terms: “dental implants,” “dental implantation,” or “dental restoration failure,” combined with “retreatment,” “replacement,” or “reoperation.” A hand search of selected journals was also performed. Of the retrieved 668 publications, 8 retrospective clinical studies met the inclusion criteria, providing the survival outcome for 673 implants in 557 patients after retreatment. Implant- and patient-related characteristics related to implant failures were assessed.
The weighted mean survival rate for implants after retreatment was 86.3%, with follow-up ranging from less than 1 year to over 5 years. The survival rates of smooth-surfaced and rough-surfaced implants were compared in 217 retreated implants, revealing a significantly higher survival rate for rough-surfaced implants than for smooth-surfaced implants (90% versus 68.7%). Insufficient data were available to evaluate the effect of patient- or treatment-related characteristics on the survival of implants after retreatment.
The survival rate of retreated implants is lower than that generally reported after initial implant placement. Higher survival rates were reported with rough-surfaced implants than with smooth-surfaced implants in retreatment. An overall implant survival rate of 86.3% after retreatment suggests that most initial implant failures are likely attributable to modifiable risk factors, such as implant architecture, anatomic site, infection, and occlusal overload.
A comparatively high survival rate of 86.3% for implant retreatments was reported with follow-up ranging from less than 1 year to over 5 years. Risk factors associated with initial implant failures, such as implant architecture, anatomic site, infection, and occlusal overload, should be assessed and modified when possible at retreatment.
Implants are widely used to support dental prostheses, consistent with their overall high survival rate. Nevertheless, the high survival rate of implants has hampered efforts to identify specific risk factors for implant failures. Early implant failures typically occur before or at the abutment connection and are generally associated with minor peri-implant bone loss. Late implant failures, in contrast, occur after occlusal loading and are often associated with extensive peri-implant bone loss. Available evidence suggests that early failure rates (0.76% to 7.47%) are somewhat lower than late (5 to 10 years) implant failure rates (2.1% to 11.3%).
Early implant failures have been commonly attributed to altered or poor wound healing, which impedes or prevents osseointegration. Additional factors, such as variation in surgical technique, poor bone quality, postoperative infection and inflammation, and occlusal overload, have been implicated in early implant failures. Late implant failures typically result from a breakdown in osseointegration, often after functional loading of the implant-supported prosthesis. Late implant failures have been generally attributed to occlusal (biomechanical) overload or peri-implantitis ; however, the biological basis for late implant failures remains largely unknown.
As with chronic periodontitis and tooth loss, implant failures show some evidence of clustering (occurrence of multiple implant failures) in selected patients and patient populations. The cluster behavior shown in one-third (33%) of patients with failed implants suggests that specific patient-related and/or treatment-related factors, such as genetic background, systemic conditions, bone quality, or occlusion, might influence implant survival. Consistent with this hypothesis, studies have demonstrated that smoking and periodontitis predict implant failures. Poorly controlled diabetes mellitus (DM) adversely affects oral wound healing and would thereby be expected to increase the risk of early implant failures and peri-implantitis.
Different treatment-related factors have been associated with an increase in implant failures. Studies have reported higher failure rates of implants placed adjacent to teeth than those of implants placed in edentulous arches. Studies have shown that early failures occurred predominantly in the posterior maxilla. Increased early failure of implants was observed for narrow-diameter implants. Larger-diameter implants have been also associated with early implant failures. Wide-diameter implants were used when primary stability could not be achieved with a standard-diameter implant or when placed in unfavorable situations such as poor bone density. Wide-diameter implants also increased the risk of buccal bone dehiscence.
A systematic review investigated the outcome of replaced implants after removal of failed implants in publications up to late 2014. An 88% survival rate was reported in this review, but due to study heterogeneity, no further quantitative analysis was available. The purpose of the present systematic review was to examine the survival rate of implants placed after implant failures from currently available studies and to determine factors that might affect the outcome of replacement.
Material and methods
The focused patient, intervention, comparison, and outcome (PICO) question was “In partially edentulous individuals, what is the survival rate of implants replacing a previously failed original implant?” Factors influencing the second implant survival were evaluated as secondary outcomes.
The following electronic databases were searched: OVID Medline, EMBASE, and SCOPUS (S.O., H.J.S.). The search strategy developed for OVID Medline was tailored for each database to account for changes in subject heading and search distinctions. The literature search used the following terms (MeSH words) with any synonyms and closed terms: “dental implants,” “dental implantation,” or “dental restoration failure,” combined with “retreatment,” “replacement,” or “reoperation.” The searches were limited to the English language, humans, and years between 1991 and 2018. Further studies were identified from a manual search of the bibliographies of relevant articles and review articles.
After the elimination of duplicate records, titles and abstracts were independently screened (S.O., H.J.S.). The reviewers agreed upon selection of 14 articles, warranting full-text access, and narrowed the final list to 8 articles based on the following inclusion criteria: prospective or retrospective clinical studies, the placement of the second implant at a previously failed site, and at least 15 partially edentulous subjects and 20 implants were included in the study. Exclusion criteria were in vitro studies, animal studies, case reports, and review articles.
The risk of bias of the selected studies was conducted by using the Newcastle-Ottawa scale for cohort studies, which was designed to assess the quality of nonrandomized studies with their selections of exposed and nonexposed cohorts, comparability based on the controls of risk factors, and completeness of outcomes. Risk of bias was assessed at the study level.
The following data were extracted from selected clinical studies: number of participants, participant demographics (age and sex) and characteristics (medical and social history), implant site location, reported etiology of initial implant failure(s), number of implant failures before retreatment at the site, follow-up period, timing of retreatment with respect to initial implant failure, bone augmentation before retreatment (yes/no), implant features (surface design, length, and diameter) at retreatment, antibiotic administration, and the type of restoration in patients with late failure.
The primary outcome in this review was the survival rate of retreated implants. The secondary outcomes were the survival rate of retreated implants with respect to implant surfaces (smooth versus rough) and implant sites (anterior versus posterior). The weighted mean survival rate was calculated based on the sample size of each study. Odds ratios were calculated to compare implant surfaces and implant sites in relation to survival after retreatment. Data analysis was performed with 2D graphing using a statistics software program (GraphPad Prism, v5; GraphPad Software, Inc).
The initial online and manual searches identified 1331 publications. After the elimination of duplicates, 668 references were available for review ( Fig. 1 ). Eight retrospective clinical cohort studies met the inclusion criteria ( Table 1 ). One clinical study was excluded because survival after retreatment was evaluated after 2 prior implant failures—exclusion was based on a concern that prior retreatment might introduce additional etiological factors influencing implant placement and outcomes. Another clinical study examined the survival rate of implants placed at the time of initial implant failure (removal) and was excluded because of the small sample size (10 participants) and the number of implant failures requiring retreatment (16 implants).
|Quality Assessment Criteria||Acceptable||Alsaadi et al||Grossman and Levin||Machtei et al||Kim et al||Mardinger et al||Wang et al||Manor et al||Chrcanovic et al|
|Representativeness of exposed cohort?||Representative of average adult in community|
|Selection of nonexposed cohort||Drawn from same community as exposed cohort|
|Ascertainment of exposure||Secured records||1||1||1||1||1||1||1||1|
|Demonstration that outcome of interest not present at start of the study||Yes||1||1||1||1||1||1||1||1|
|Study controls for age/sex||Yes|
|Study controls for at least 3 additional risk factors?||Smoking, diabetes, periodontitis||1|
|Assessment of outcome||Record linkage||1||1||1||1||1||1||1||1|
|Follow-up long enough||Follow-up (>3 y)||1||1||1||1|
|Adequacy of follow-up||Small number of subject loss||1||1||1||1||1||1||1||1|
|Overall quality score (maximum=9)||4||4||4||5||5||5||5||5|
The quality assessment of the selected studies is summarized in Table 1 . The selected studies exhibited a moderate risk of bias based on their overall quality scores (4.6 ±0.52).
Participant characteristics among the selected studies are summarized in Table 2 . Clinical characteristics and outcomes from the selected studies are summarized in Table 3 . Two studies allowed for a comparison of survival rates after retreatment based on the implant surface. Four studies allowed for a comparison of survival rates after retreatment based on the location where implants were replaced.
|Study||Number of Implants||Sex||Age (Mean/Range)||Patient Characteristics||Timing of Initial Failure (Before or After Loading)||Attributable Causes of Failure|
|Alsaadi et al||58||23 men; 18 women||44 y/24-84 y||12 smokers; 2 osteoporosis; 2 hyperthyroid; 4 hypothyroid||Not specified||No specified reasons for implant failures|
|Grossman and Levin||31||25 men; 3 women||33.2 y/20-53 y||3 smokers; 1 diabetes; 16 periodontitis||Preloading (100%)||No specified reasons for early implant failures|
|Machtei et al||79||56 patients (sex not specified)||53.7 y/22-84 y||15 smokers; 3 diabetes||Preloading (90%); postloading (10%)||Inflammation and suppuration (23%), prolonged acute pain (7%), no specified reasons for mobility of implants (70%)|
|Kim et al||60||35 men; 14 women||53.2 y||0 smokers; 0 patient with systemic disease||Preloading (68%); postloading (32%)||Inflammation (5%), infection (5%), fixture fracture (1.7%), malposition (1.7%), no specified reasons for lack of osseointegration (87%)|
|Mardinger et al||144||60 men; 84 women||56.5 y/31-73 y||34 smokers; 0 patient with systemic disease||Postloading (100%)||Infection (15%), overload (17%), unknown (31.3%), no specified reasons for lack of osseointegration (38%)|
|Wang et al||67||38 men; 28 women||42.3 y/21-68 y||3 diabetes; 12 smokers||Preloading (100%)||No specified reasons for early implant failures|
|Manor et al||75||34 men; 41 women||56.5 y/32-76 y||19 smokers||<12 mo after insertion (77.3%)||Overload, peri-implantitis, lack of osseointegration from infection (77.3%)|
|Chrcanovic et al||159||47 men; 51 women||54.2 y||6 diabetes; 26 smokers; 3 hypothyroidisms; 21 hypertensions; 17 antithrombotic; 12 antidepressants||At the second stage (26.1%); <12 mo after insertion (49%)||Implant fractures; no specified reasons for loss of osseointegration|