Evaluation of 3- to 8-Year Treatment Outcomes and Success Rates With 6 Implant Brands in Partially Edentulous Patients

The science of implantology is progressing rapidly, with new systems being presented every year. The replacement of missing teeth with dental implants has become an acceptable standard of care over the past several decades.^1,2  Functional and esthetic physiological outcomes and psychological benefits have been reported for the use of implant-supported fixed partial dentures (FPDs) in patients who object to removable partial denture (RPD) use.^3–5  The success rate of implant prosthetics is higher than that of natural teeth–supported traditional prostheses.^4,5  Compared with tooth-supported cantilever FPDs, implant-supported FPDs for the treatment of distal extension base situations are less prone to problems.^3–5  Different risk factors can negatively affect implant longevity. Anatomic features, mastication dynamics, and adequate implant selection are all significant for long-term implant prognosis.⁶ 

Implant design features such as macro- and micro-design may influence overall implant success. Limited information is currently available. Appraising the current literature on this subject and combining existing data to verify the presence of any association between the selected characteristics may be critical in the achievement of overall implant success.⁷  Research on new macrostructures and nanomorphology should result in a better qualitative and quantitative osseointegration response, with a predictability of the clinical results and long-term success of the implants.⁸ 

Implants with rough surfaces showed a statistically higher survival rate than machined implants at all intervals. The prosthetic design, veneering material, and the number of prostheses per arch had no influence on the prosthodontic survival rate. Implant number and distribution along the edentulous maxilla seemed to influence the prosthodontic survival rate.⁹ 

Because the types of implant systems with different macro and microstructure might affect the failure or success of treatment, we performed clinical and radiological evaluations to estimate the success rate of different implant systems in partially edentulous patients after a functional period of 3–8 years.

This retrospective study was performed in the Mashhad Department of Implantology, School of Dentistry, and Mashhad Dental Research Center, Mashhad University of Medical Sciences, Mashhad, Iran. Patients provided written informed consent prior to participation, and information was given to each patient regarding alternative treatment options. All patients were partially edentulous. Inclusion criteria were the presence of natural teeth adjacent to the edentulous space and opposing natural teeth or FPD. Exclusion criteria were poorly controlled diabetes, recent history of coronary heart disease, severe bleeding disorders, immediately loaded implants, any type of prosthesis other than single crowns, or prosthesis complications such as metal or porcelain fractures. Because smoking may affect implant-related surgical procedures, smokers were also excluded.

A total of 168 patients with 550 implants were selected from the archive of Mashhad Department of Implantology. Eighty-eight patients (mean age, 52 years) with 268 implants participated in this 5-year retrospective study. Twenty-five patients were lost during follow-up: 4 patients died, 12 patients were too ill to participate in the evaluation, and 9 patients had moved. Another 55 patients were excluded because of exclusion criteria.

Patients had different implant brands, including 110 BioHorizons (BioHorizons implant, BioHorizons Inc, Birmingham, Ala), 60 ITI (ITI implant, Straumann Inc, Basel, Switzerland), 55 Paragon (Paragon [Swissplus] implant, Zimmer Dental Inc, Carlsbad, Calif), 18 Xive (Xive implant, Densply Friadent International Inc, Mannheim, Germany), six 3i (3i implant, Biomet 3i Inc, Palm Beach Gardens, Fla), and 19 Allfit (Allfit implant, IHDE Dental implant Inc, Eching, Germany). All implants were at least 9 mm long and had a diameter of at least 3.3 mm, based on the available bone volume. After selecting the sample size, patients were recalled to the department and data were registered by clinical and radiological evaluation using a specified check list. Information included complete medical and dental history, smoking habits, and clinical and radiographic findings.

A 1-stage surgical approach was performed on 60 implants. All other implants were healed with a 2-stage surgical approach. Implants were permitted to heal for at least 3 months prior to prosthesis fabrication. During the osseointegration period, patients wore interim RPDs or previous FPDs were modified as provisional restorations and cemented with provisional cement (Temp-Bond, Dentsply Caulk, Milford, Del) on the previously prepared teeth. In nonesthetic critical areas, no interim prosthetic treatment was performed during the osseointegration period. Single crowns were fabricated and cemented with glass ionomer cement (76-1 Hasunumy-Cho, Itabashi-Ku, Tokyo, Japan).

Standardized dental radiographs were taken at baseline, 12 weeks, 1 year, and at the time of study to measure the amount of peri-implant bone over the study period. The same film type (Kodak Ultra-speed DF-58, Eastman Kodak Co, Rochester, NY) was used for all patients. Radiographs were exposed at 60 kvp, 15 mA, and 0/5 seconds and were developed under standardized conditions (810Plus, Velopex Intl Inc, St Cloud, Fla).

All radiographs were evaluated by a blinded dentist trained in evaluating panoramics. To calculate the measurement error, 4 study films (each containing 5 radiographic study sites) were randomly selected. The examiner evaluated the first 5 radiographic sites and then proceeded to the next 5 sites. This procedure was repeated for a total of 3 times, with a new randomized order each time.

Peri-implant tissues were examined using the following clinical parameters: bleeding tendency with the sulcular bleeding index, peri-implant probing depth (PPD), and peri-implant marginal bone loss (BL). The bleeding on probing (BOP) was determined after the insertion and movement of a standard periodontal probe into the peri-implant sulcus parallel to abutment. A dichotomous scale was used: score 0 (no bleeding) or 1 (with bleeding). The sulcus probing depth was defined as the distance from the gingival margin to the bottom of the sulcus pocket. The sulcular probing depth (in mm) was measured with a periodontal probe (Williams, Hu-Friedy, Chicago, Ill) with a light force of ∼0.2 N at 4 sites around the implant (mesial, distal, buccal, and lingual). The BL was defined as the distance (in mm) between the implant shoulder and the first visible bone contact (ie, a higher contact in mesial or distal counted as BL).¹⁰  Possible distortions of the BL on the radiograph were assessed by comparing the actual and radiographic implant lengths.¹¹ 

We used the modified Albrektsston criteria (1989) to assess implant success.¹²  An implant was considered to have failed when one of the following conditions was present: (1) peri-implant radiolucency, (2) any sign of mobility, or (3) signs or symptoms of pain or infection around the implant.^12,13  Implant mobility was tested using the handles of 2 dental mirrors.

Relationships were tested between clinical measures and the implant brand using analysis of variance (ANOVA).

Two variables were compared using a post hoc test. An analysis of covariance was used to evaluate the effect of time on the other variables. Possible relationships between length, width, and time with BL were tested using logistic regression analysis. All statistical analyses were performed using SPSS 11 (SPSS, Chicago, Ill). A P value of <.05 was considered statistically significant.

Unilateral 1-way ANOVA with post hoc analysis revealed significant differences:

1. In BL, between ITI and Xive (P = .031), between BioHorizons and Allfit (P = .015), between Paragon and Xive (P = .040), and between Allfit and Xive (P = .001).

2. In BOP, between BioHorizons with Xive (P = .011), Allfit (P = .002), ITI (P = .006), or 3i (P = .008), between Xive with Paragon (P = .006), between ITI with Paragon (P = .003), between Paragon with 3i (P = .005), Allfit (P = .001).

3. In PPD, between Allfit with ITI (P = .023), BioHorizons (P = .007), or Paragon (P = .023), Xive (P = .003).

Minimal and maximal BL and PPD values were seen in Allfit and Xive, respectively (Figures 1 and 2). Minimal and maximal BOP values were seen in the 3i and Paragon systems, respectively (Figure 3). Figures. 4, 5, and 6 show differences in BL, PPD, and BOP between different jaw locations, respectively. The PPD and BL values were greater in the anterior region than in the posterior or premolar area (Figures 4 and 5). The BOP was greater in the posterior than in the anterior region (Figure 6).

Pearson correlation analysis revealed a significant relationship between the BL and PPD; between PPD and BL, BOP, or time; and between BOP and PPD or time (Table 1). No significant relationship was seen between the implant type and BL or PPD when time was considered as a covariate (Tables 2 and 3).

The type of jaw (mandible or maxilla) had no effect on the clinical variables (Table 4). When time, location, and jaw were considered together, a significant relationship in PPD was observed with location or time (Table 5). The mean PPD in the mandible was insignificantly more than that in the maxilla (Table 6). Logistic regression analysis revealed no significant relationship in the BL with the implant length, implant width, or time (Table 7). Table 8 shows the failure of the different implant systems.

The aim of this study was to evaluate treatment outcomes 3 to 8 years after implantation of 6 different implant systems. Although there were significant differences between the systems in BL and PPD, these differences disappeared when we considered time as a covariate (Tables 2 and 3). Similarly, Ozkan et al¹⁴  previously found no significant differences in marginal BL values among 3 implant brands, all of which exhibited similar positive treatment outcomes after 3 years.

As expected from their roles as indicators of peri-implantitis,¹⁵  we observed a significant relationship between PPD and BL, which implied that PPD is a good predictor for peri-implant BL. A 2-year longitudinal study¹⁶  found that the probing attachment level and radiographic parameters together serve as a good predictor of peri-implant tissue status. In contrast, previous long-term evaluations^11,17  found no relationship between the clinical parameters and BL by radiography. A 5-year study by Weber et al¹⁸  found a very low rate of correlation between individual clinical parameters and radiographically measured BL, suggesting that these parameters are of limited clinical value in assessing peri-implant BL.

The BOP, PPD, and BL values differed between the different systems, perhaps because of differences in the experience of the surgeons or the surgical procedure used. Melo et al¹⁹  found no survival differences among implants placed by surgeons with differing education levels. In the present study, 3 surgeons with 5 years of surgical experience performed all surgical procedures. Although submerged placement of the implant fixture during the healing period was long considered to be a prerequisite for osseointegration, Schroeder et al confirmed the predictability of the nonsubmerged fixure.^20,21  Recent investigations in partially edentulous patients have revealed no difference in success between submerged and nonsubmerged implants.^22–24  In the present study, favorable results were obtained with both submerged and nonsubmerged implants.

Maximal and minimal BL and PPD values were seen in the anterior and premolar regions, respectively (Figures 4 and 5). Surgeons typically hide the metal collar of a prosthesis or implant by deeply inserting the collar into the anterior region. This increases future PPD measurements; on the other hand, the anterior area also permits easier dental hygiene. Thus, the BOP index was smaller in the anterior region than in other areas (Figure 6). The mean value of PPD in the mandible was more than in the maxilla, but this difference was not significant (Table 6). Surgeons try to insert implants deeply in the anterior mandible to prevent thread exposure on the buccal or lingual sides as a results of a narrow ridge in this area. A significant relationship was also seen between the effect of location and time on PPD. No association was found between the implant length and peri-implant BL (Table 7). Previous studies have shown that shorter or wider implants, implants supporting a fixed prosthesis, and implants in smokers are associated with greater BL.²⁵  Indeed, implant length is reported to be the most significant factor in dental implant maintanence.²⁶  When used appropriately, implants of 6 to 9 mm in length have a cumulative survival under function comparable with longer implants.²⁷  Within the limits of their study design and observation period, Romeo et al found that implant size did not compromise the effectiveness of implant therapy.²⁸  Because of anatomical limitations, posterior-placed implants are often shorter than those inserted in the anterior regions and thus are associated with a poorer success rate in both the mandible and maxilla.²⁹ 

The success rate of the inserted implants was 96.27%, which is comparable with other long-term follow-up studies.^15,30,31  One ITI and 4 BioHorizons failed to osseointegrate (“early failure”) and had to be removed after 1 month. In addition, 1 ITI and 2 BioHorizon implants had to be removed after 3 years. The inability to maintain the achieved osseointegration under functional conditions may be considered a late loss.³⁰ 

The lack of standardized and internationally recognized success criteria makes it difficult to compare different studies. In fact, some authors have defined their own success criteria.^15,31  Ferringno et al¹¹  and Buser et al³²  used the criteria proposed by Buser et al,¹⁰  while Brocard et al³¹  used a combination of the Albrektsson et al¹²  and Buser et al¹⁰  criteria. In the present study, we used the modification of criteria proposed by Albrektsson et al^13,14  in 1986. Although it is difficult to compare the present results to others because of these differences in criteria, our results seem to agree with results from comparable studies.^10,11,15,31 

One of the disadvantages of the conventional radiograph system is that it reveals only 2-dimensional information and potentially overlooks tissue breakdown on the buccal and lingual peri-implant sites. Different methods of measuring bone levels adjacent to dental implants have been developed.^33–35  Although such systems would be useful for standardized scientific research, they might be too difficult for use in a clinical setting. Based on a recent systematic review, there is no evidence showing that any particular type of dental implant has superior long-term success. These findings are based on a few randomized controlled trials, often at high risk of bias, with few participants and relatively short follow-up periods.³⁶  The results of our study showed that implant systems provided stable bases for long-term prosthesis support. No significant differences were seen between different systems. It is important to monitor the peri-implant bone level by taking annual radiographs and measuring probing depth, especially after prolonged service.

BL

bone loss

BOP

bleeding on probing

FPD

fixed partial dentures

PPD

peri-implant probing depth

RPD

removable partial denture

The authors would like to thank the research Council of Mashhad University of Medical Sciences for financial support of this project.