PDF
Citation
Print
Introduction
Reconstruction of fully or partially edentulous areas with osseointegrated implants is one of the most reliable prosthetic treatments, with survival rates above 90% in long-term studies.1,2 Despite these high survival rates, there are still complications and failures in implant treatment. Previous studies have reported early failures in 0.7% to 7.4% of cases and late failures in 2.1% to 11.3% of cases.1 Therefore, the identification of risk factors associated with implant failure is essential for treatment planning.
Risk factors for adverse outcomes range from implant design to coexisting systemic disease.3 In several studies, diabetes, steroid therapy, osteoporosis, chemotherapy, and head and neck irradiation have been considered to be contraindications for the placement of implants.4 Conversely, in other studies, individual medical problems did not correlate with an increase in implant failures. Rather, implant success was affected by bone quantity and quality and by surgical techniques.5 A variety of studies have reported that tobacco use has a negative effect on implant survival.6,7 However, no obvious evidence has been presented concerning the influence of tobacco use on the survival rates of implants.2,7 The effect of a history of periodontal disease on implant failure is also controversial. Several authors reported that there may be an increased risk for implant failure in periodontally compromised patients,8,9 while other authors insisted that the presence of periodontal pathogens at peri-implant sites does not necessarily indicate future implant failure.10,11
Because the results obtained from numerous studies are contradictory, the questions raised by patients cannot be answered. In spite of the wide range of available studies, consistent conclusions cannot yet be drawn on the relationship between these factors and the long-term outcomes of implants. There is a clear need to obtain further scientific evidence in this area. Therefore, the aim of the present study was to evaluate the long-term survival of implants retrospectively and determine the risk factors associated with implant failure.
Materials and Methods
Data collection
A retrospective chart review was conducted for all implants that were placed at the Department of Peri-odontology of the Dental Hospital of Gangneung-Wonju National University from January 1998 to December 2012 (IRB 2013-13). Of all implants that were included the chart review, only implants that were followed up until June 2013 were included in this study. Data were collected from clinical and radiographic examinations from previous visits. The information gathered included the following:
Gender
Age (classified into three categories: < 50 years, 50 - 59 years and ≧ 60 years)
Smoking status (people smoking 1 cigarette or more per day at the time of implant placement were categorized as a smokers)
Implant diameter (classified into three categories: < 3.75 mm, 3.75 - 4.5 mm, ≧ 5 mm)
Implant length (classified into two categories: < 10 mm, ≧ 10 mm, the classification criteria of 10 mm is based on the definition of a short implant12)
Surface of implant (classified into three categories: blasted, sand-blasted and acid etched, and anodized surface)
Location of implant within the dental arch (classified into four categories: maxillary anterior, maxillary posterior, mandibular anterior and mandibular posterior regions)
Surgical techniques (specific procedures that were performed to place the implants: guided bone re-generation, sinus elevation with a crestal approach and sinus elevation with a lateral approach)
Complications (biological and mechanical complications: peri-implant mucositis, peri-implantitis, implant periapical lesion and fracture of fixture, screw or prosthesis)
Dates of implant placement, prosthesis placement, implant removal and the last follow-up visit
An implant failure was defined as the removal of the implant for any reason.13 Implants were regarded as surviving when they were present in the oral cavity at follow-up visits. Total survival time was defined as the period from the date of implant placement to the date of implant removal or the last follow-up visit.13 Implant failure was classified as early failure if it occurred before functional loading, or late failure if it occurred after functional loading.14
Statistical analyses
After the data collection, statistical analyses were performed. The cumulative survival rates of implants were calculated by the life table method at annual intervals and a Kaplan-Meier survival curve was obtained. To determine the potential risk factors associated with implant failure, a simple logistic regression analysis was performed. Implant failure was the dependent variable and all independent variables were entered separately as categorical variables. Following the simple regression analysis, independent variables with P < 0.1 were selected and included in a multiple regression analysis to eliminate possible confounding variables. A multiple logistic regression analysis was performed for implant failure as the dependent variable and independent variables with P < 0.05 were considered statistically associated with implant failure. In all types of analyses, each implant was regarded as the analysis unit. Thus multiple implants from the same patient had the same patient-related covariates. All analyses were conducted using statistical software (SPSS™ 20, IBM Inc., Chicago, IL, USA).
Results
During the period from January 1998 to December 2012, 5019 implants were placed in 1921 patients. Of these patients, 1146 were excluded from this study because of a lack of follow-up visits. Thus the study group was composed of 2265 implants from 775 patients.
Demographic characteristics
The study group consisted of 296 females and 479 males. The mean age at the last visit was 55.5 years (range: 19 to 81 years). Of 2265 implants, 854 implants (37.7%) were placed in females, and 1411 implants (62.3%) were placed in males. Six hundred thirty-four implants (28.0%) were placed in smokers, and 1631 implants (72.0%) were placed in non-smokers (Table 1).
Table 1
Distribution of implants according to demographic characteristics
Implant characteristics
The majority of the implants were of a diameter ≧ 3.75 mm (95.3%), and 4.7% were of a diameter < 3.75 mm. Most of the implants were ≧ 10 mm (95.8%) in length. The implants had a variety of surfaces including blasted (49.2%), anodized (30.7%), and SLA surfaces (20.1%) (Table 2).
Table 2
Distribution of implants according to implant characteristics
Surgical characteristics and the existence of complications
One thousand one hundred and twenty-five implants were inserted in the maxilla (195 in the anterior and 930 in the posterior) and 1140 in the mandible (75 in the anterior and 1065 in the posterior). The guided bone regeneration (GBR) technique was used in 559 implants, and 368 implants were placed after maxillary sinus elevation (198 with the crestal approach and 170 with the lateral approach). The others (1338 implants) were placed without any specific procedure. Postoperative complications including peri-implant mucositis, peri-implantitis, implant periapical lesions and fractures of the fixture, screw or prosthesis occurred in 354 implants (15.6%) (Table 3).
Table 3
Distribution of the implants according to surgical characteristics and the existence of complications
Implant survival
Of the 2265 inserted implants, 55 (2.4%) failed. Twenty-five implants (1.1%) failed before loading (early) and 30 (1.3%) failed after loading (late). According to the Kaplan-Meier lifetime analysis (Table 4, Fig. 1), the cumulative survival rate after 5 years of loading was 97.2%, and after 15 years of loading was 95.2%.
Table 4
Cumulative survival rate of inserted implants
Risk factors associated with implant failure
All independent variables including gender, age, smoking status, implant diameter, implant length, surface of implant, location of implant, surgical technique and existence of complications were entered into a simple logistic regression analysis (Table 5 - 7). Among these variables, gender (P = 0.016), smoking status (P = 0.001), location of implant (P = 0.020) and existence of complications (P = 0.002) were statistically associated with implant failure and included in the multiple regression analysis.
Table 5
Simple logistic regression analysis for risk factors associated with implant failure (Demographic variables)
| Odds ratio | 95% Confidence Interval | P value | ||
|---|---|---|---|---|
| Gender | 0.016* | |||
| Female (baseline) | ||||
| Male | 2.206 | 1.156 to 4.206 | 0.016 | |
| Age (yr) | 0.526 | |||
| < 50 (baseline) | ||||
| 50 - 59 | 0.757 | 0.385 to 1.487 | 0.419 | |
| ≥ 60 | 0.654 | 0.311 to 1.373 | 0.262 | |
| Smoking | 0.001* | |||
| No (baseline) | ||||
| Yes | 2.547 | 1.489 to 4.356 | 0.001 | |
Table 6
Simple logistic regression analysis for risk factors associated with implant failure (Implant variables)
Table 7
Simple logistic regression analysis for risk factors associated with implant failure (surgical variables and complications)
As a result of multiple logistic regression analysis (Table 8), the variables statistically associated with implant failure (P < 0.05) were smoking status (P = 0.049) and existence of complications (P < 0.001).
Table 8
Multiple logistic regression analysis for risk factors associated with implant failure
| Variables | Odds ratio | 95% Confidence interval | P value |
|---|---|---|---|
| Gender | 0.272 | ||
| Female (baseline) | |||
| Male | 1.501 | 0.727 to 3.098 | 0.272 |
| Smoking | 0.049* | ||
| No (baseline) | |||
| Yes | 1.839 | 1.002 to 3.377 | 0.049 |
| Location | 0.085 | ||
| Maxillary anterior (baseline) | |||
| Maxillary posterior | 2.395 | 0.722 to 7.943 | 0.153 |
| Mandibular anterior | 1.895 | 0.306 to 11.726 | 0.492 |
| Mandibular posterior | 1.166 | 0.336 to 4.055 | 0.809 |
| Complications | < 0.001* | ||
| Absence (baseline) | |||
| Presence | 2.977 | 1.685 to 5.258 | < 0.001 |
The implants placed in smokers were 1.839 times more likely to fail than the implants placed in non-smokers. The implants with complications were also at a greater risk of failure compared to the implants without complications, with an odds ratio of 2.977 : 1.
Discussion
The purpose of this retrospective study was to analyze the long-term outcomes of 2265 implants and determine the risk factors associated with implant failure. A retrospective chart review of implants placed at the department of periodontology of the dental hospital of Gangneung-Wonju national university from January 1998 to December 2012 was performed.
In the present study, the cumulative survival rate up to 15 years was 95.2%. Previously, Simonis et al.15 reported a long-term cumulative survival rate of 82.94% up to 16 years. Additionally, Roos-Jansåker et al.2 assessed the long-term outcome of implant therapy and reported that the overall survival rate up to 14 years was 95.7%. Thus, the outcome of the present study is consistent with previous studies.
In this study, only 55 implants were removed. Some of these were concentrated in a small number of individuals. Fifty-five implant removals occurred in 47 patients and 15 of 55 removed implants were in 7 patients. Among them, five were smokers and two were non-smokers. Peri-implantitis occurred in 2 patients and implant periapical lesion occurred in 1 patient.
To identify the risk factors associated with implant failure, statistical analyses were performed. Using multiple logistic regression analysis, a significant relationship was found between implant failure and independent variables including smoking status and existence of complications. Smoking status has previously been regarded as a risk factor for implant failure. Bain and Moy16 reported a failure rate of 11.3% in smokers and 4.76% in non-smokers. In more recent study, Huynh-Ba et al.17 studied 273 implants placed in the posterior maxilla and reported that smoking markedly increased the risk for implant failure. In terms of existence of complications, Simonis et al.15 referred to peri-implantitis as a complication with a very high risk of implant loss; of 21 cases, only 5 implants were functional after 16 years despite comprehensive treatments including implant debridement and administration of systemic antibiotics.
Some variables that were not significantly related to implant failure in multiple regression analysis but reached statistical significance in simple regression analysis included gender and location of implant. Males (odds ratio = 1.501) and maxillary posterior implants (odds ratio = 2.395) were associated with implant failure. These variables were previously identified as risk factors for implant failure. Zupnik et al.18 reported that gender was the parameter with the strongest correlation with implant failure in a meta-analysis that demonstrated that males had a higher prevalence of destructive periodontal disease.19 Because periodontal disease is widely recognized as a risk factor for peri-implantitis and implant failure,20,21 the higher prevalence of periodontal disease in males may explain the outcome of the study. Anitua et al.22 reported significantly lower survival for implants placed in the maxilla compared to the mandible. This difference might be the outcome of the less favorable osseous situation in patients requiring maxillary reconstruction. In addition, Moy et al.3 studied 4680 implants in 1140 patients and reported that implants placed in the maxilla failed at almost twice rate of those placed in the mandible.
The present study had some limitations. First, an inherent limitation of retrospective studies is a risk for bias because only the implants that were followed up were included in the study group. The dropout rate (1146 of 1921 patients, 59.7%) could be considered high and must be considered when interpreting the results. Previous studies have suggested this as a reason for incomplete results.17,23 Second, because the number of failed implants was very low (55 of 2265 implants, 2.4%), definitive conclusions could not be drawn. Alsaadi et al.24 reported that due to the low number of implant failures, only potentially influential factors could be identified. Finally, numerous variables that are known to be significantly associated with implant failure were not included in our analyses. For example, diabetes has been significantly correlated with implant failure.18 Rodrigo et al.25 reported that resonance frequency analysis (RFA) values at restoration placement could significantly predict the outcomes of implants. Quirynen et al.26 reported that a large proportion of failing implants may be explained by the lack of proper supportive periodontal therapy (SPT), and Ong et al.27 proposed that the irregularity of SPT might influence the outcomes and differences between studies. Therefore, well-designed, controlled prospective studies that consider the above limitations are required.
XML Download