Factors Influencing Resonance Frequency Analysis (RFA) Measurements and 5-Year Survival of Neoss Dental Implants

Background Diagnostic instruments based on resonance frequency analysis (RFA) can be utilised to assess dental implant stability during treatment and follow-up. Aim The aim of the present study was to investigate the influence of patient- and implant-related factors on implant stability and the 5-year implant survival. In addition, the influence of stability (ISQ value) at placement and abutment connection on implant survival was evaluated. Materials and Methods RFA measurements from a total of 334 consecutive patients with 745 dental implants (Neoss Ltd., Harrogate, UK) were retrospectively analysed after at least 5 years in function. Statistics were used to evaluate the influence of the different variables on implant stability and implant survival. Odds ratio calculations were performed to compare the risk for implant failure using 60, 65, 70, and 75 ISQ as threshold levels at placement and loading. Results A total of 20 implant failures in 14 patients were noted during the 5 years of follow-up, giving an overall cumulative survival rate (CSR) of 97.3% at the implant level and 95.8% at the patient level. Gender, jaw, position, bone quality, and implant diameter had an influence on implant stability at placement. Jaw, bone quality, and implant diameter had an influence on stability after 3-4 months of healing. More failures were observed in full than in partial rehabilitations. Age, gender, jaw, position, bone quantity, bone quality, implant diameter, and implant length had no influence on implant survival. Implants with ISQ values below the threshold levels showed lower survival rates compared to implants with values above these levels. Conclusions The present study showed a significantly higher risk for implant failure, showing an ISQ value below 70 and 75 at placement or after 3-4 months of healing. The results indicate that RFA measurements can be used to identify implants with increased risk for failure.


Introduction
Firm primary implant stability is regarded as a determinant of success in implant dentistry [1]. is notion is based on early observations that implant failure is more common in the presence of compromising biomechanical factors, such as in case of low bone density, short implants, grafted bone, premature loading, and overloading [2,3]. One explanation is that low implant stability may lead to micromotions at the bone-implant interface during healing and loading, which in turn stimulates to fibrous soft tissue formation rather than bone healing. Diagnostic instruments based on resonance frequency analysis (RFA) are available for assessment of implant stability as a function of interface stiffness during treatment and follow-up [4]. e technique measures the resonance frequency (RF) of a transducer that is attached to the implant. With the existing commercial instruments, the RF in Hz is translated to an implant stability quotient (ISQ) unit between 1 and 100, where increased values reflect increased stability [5]. Experimental research using displacement measurements has shown a correlation between RFA and micromobility when applying a lateral load to the dental implant [6]. e bone density at the implant site seems to be the most important determinant of primary stability [6][7][8][9], although factors such as implant design, length, and diameter seem to affect RFA measurements as well [10,11]. Experimental and clinical studies have demonstrated an increase of ISQ units with time as a consequence of bone formation and maturation at the implant interface [12][13][14]. According to the manufacturers and some authors, the technique can be used to measure that sufficient stability has been achieved prior to as well as identifying implants with low stability in order to take measures to avoid implant failure [4]. Although clinical studies have shown implants with low and/or falling ISQ units over time to be more prone to failure [10,15,16], the ability of the RFA technique to predict implant failure has been questioned by several authors [17][18][19].
e aim of the present study was to retrospectively investigate the influence of patient-(age, gender, jaw, indication, bone density, and bone volume) and implantrelated (diameter and length) factors on implant stability as assessed by RFA measurements and the 5-year implant survival in 334 consecutive implant patients. In addition, the influence of stability (ISQ value) at placement and abutment connection on implant survival was evaluated.

Patients.
Data related to patient, type of treatment, implants, bone conditions, and outcomes at annual checkups were extracted from a simple computerised system (MS Excel, Microsoft, Redmond, USA) used to keep track on consecutive implant treatments at the three centres participating in the study. A total of 334 consecutive patients (55.4% female, 44.6% male, mean age 52 ± 11.8 years) who met the following inclusion criterion were included in the present retrospective analysis: implant treatment for replacement of one or several teeth using a one-or two-stage procedure with 3 to 4 months of healing to prosthetic loading and follow-up for at least 5 years in function (Table 1). Immediate/early loading cases were not included. All patients had been given their written consent to the treatment plan and follow-up according to the routine procedures at the centres. e study followed the directives given by the Ethical Committee at the Feltre Hospital, Feltre, Italy, and in accordance with the World Medical Association Declaration of Helsinki. e patients had received 745 dental implants (Neoss Implant system, Neoss Ltd., Harrogate, UK) in diameters from 3.5 to 5.5 mm and in lengths from 7 to 17 mm in all tooth positions in both jaws ( Figure 1, Table 2). e implant has a slightly tapered geometry (one degree) and a modified surface obtained by double particle blasting (Bimodal surface). e roughness is higher on the body and less on the neck. e implant system uses an internal flat-to-flat connection with the prosthetic parts. Bone quantity and quality had been evaluated according to the Lekholm and Zarb index [20] (Table 3). e majority of implants had been measured with RFA at placement (n � 655) and at abutment connection or prosthetic loading (n � 529) 3-6 months after implant placement by using an Osstell Mentor ™ machine (Osstell AB, Gothenburg, Sweden) in implant stability quotient values (ISQ) (Figure 2). e implants were supporting 165 single crowns, 180 fixed partial bridges, 19 fixed full bridges, and 16 overdentures (Tables 4 and 5). e prostheses were made on Neolink ™ abutments (Neoss Ltd., Harrogate, UK) made of titanium or gold depending on the material of the framework. Both porcelain and acrylic teeth were used in the study. At two of the centres, most prostheses were screw-retained and the access holes covered with composite fillings. Individual titanium abutments for cementation were also used in some cases to correct implant angulation. At the third centre, only cemented prosthetics had been used. e patients had been attending annual check-ups and any adverse events such as implant failure, soft tissue problems, or technical complications were noted in the patient charts. In this study, only implant failure was used as a parameter.   e variables tested were age, gender, jaw, indication (single tooth, partial jaw, and full jaw), bone quantity, bone quality, implant diameter, and implant length. In addition, the influence of the ISQ value at placement and abutment connection on implant survival was evaluated.

Results
Five-year survival data could be extracted from 328 of the 334 patients. Six patients with 13 implants had dropped out because of death (n � 2) or they had moved (n � 4).
A total of 20 implant failures in 14 patients were noted during the 5 years of follow-up giving an overall cumulative survival rate (CSR) of 97.3% at the implant level and 95.8% at the patient level. Eight (1.1%) were early failures (before one year of loading), and 12 (1.6%) were late failures (after one year of loading) ( Table 6).
Statistical analyses of the influence of different factors on implant survival showed more failures in full than in partial rehabilitations (p � 0.003). Age, gender, jaw, position, bone quantity, bone quality, implant diameter, and implant length had no influence on implant survival (Table 9).
In general, implants with ISQ values below the threshold levels of 60, 65, 70, and 75 ISQ at placement and loading showed lower survival rates compared to implants with values above these levels (Table 10, Figure 3). Moreover, the failure rate decreased with increasing threshold level for implants with ISQ levels ≥ the threshold,     while this was not observed for implants below the threshold levels ( Figure 3). e differences were statistically significant for threshold values of 70 and 75 ISQ at both placement and loading. When analysing ISQ values from implant placement surgery, the risk for implants failure increased to 6.3 times (p � 0.003) and 17.9 times (p � 0.001) for implants below 70 and 75 ISQ, respectively (Table 10).

Discussion
e present retrospective study showed good clinical outcome with the Neoss implant system after at least five years of loading. e CSR of 97.3 % corresponds well to the 5-year results from other studies on the same [21] and other modern implant systems [22]. In an analysis of data from the Swedish Social Insurance Agency, Derks et al. reported a   [23]. e analysis of factors influencing the RFA measurements confirmed in large what have been found and reviewed in other publications [4,5]. ere was a relation between bone density and ISQ values, which in turn explains the difference seen between the maxilla and the mandible. Implant length had no influence on the stability in the present study, which is in line with some studies [24][25][26], although other authors have found a correlation [10]. However, 4.5 mm wide implants were significantly more stable than 4.0 mm ones, which corroborates with the study byÖstman et al. [27]. Studies have demonstrated a correlation between marginal bone thickness and ISQ measurements [28][29][30], which can explain the lack of influence of implant length and the positive effect of implant diameter in the present study. Interestingly, significantly lower ISQ values were found in women than in men. is has also been noted in other studies [11,31,32] and may be related to the higher incidence of osteoporosis in women.
Clinical follow-up studies on the first generation of dental implants with a machined surface have shown higher failure rates for implants placed in compromised bone situations [1]. is includes short implants placed in small bone volumes [33] as well as implants in sites with low bone density [2,3]. In the present study using a surface-modified implant, 207 of the 745 implants were 7 mm or 9 mm long and no difference in failure rate compared to the longer implants was found. Neither did bone density had an effect on implant failure. is is in line with studies on other modern implants [33] and underlines the importance of implant surface modification for good integration [34]. Reviews of the literature have shown improved overall outcomes with surface modified implants compared with machined surfaced implants and particularly in compromised bone situations [35][36][37].  (1) Kruskal-Wallis test; (2) Mann-Whitney U-test; (3) pairwise Mann-Whitney U-tests (Bonferroni-corrected significance levels).

International Journal of Dentistry
Interestingly, implants with low stability showed lower survival rates than implants with high stability when using 60, 65, 70, and 75 ISQ as threshold levels at placement. e similar was seen at loading but with one exception. Moreover, it was clear that the failure rate for implants above each threshold level decreased in a linear fashion with increased threshold levels. However, the opposite, i.e., more failures with decreasing threshold values for implants below these levels, could not be observed. In fact, significant differences were only seen for the 70 and 75 ISQ threshold levels. e lack of differences for the 60 and 65 ISQ threshold levels is explained by the fact that too few values were available in the groups with lower values, which hampered the statistical analyses. Moreover, it is possible that some failures were due to factors not related to stability, such as infection. Nevertheless, the risk for implant failure was 3.9 times higher Pearson chi-square; (2) Fisher's exact test; (3) pairwise Fisher's exact tests (Bonferroni-corrected significance levels). 6 International Journal of Dentistry with ISQ values below than above 70 at implant placement and 6.9 times higher when using 75 ISQ as a threshold value. is is in line with the idea that implant micromotion increases the risk for soft tissue integration during initial healing, which is the reason most implant surgeons strive for high primary implant stability [38]. Moreover, it is logical that implants with weak bone integration at prosthesis connection pose a higher risk for failure than well-integrated implants, which was shown by the 6.3 times higher risk for implants with values below 70 ISQ at abutment/prosthesis connection and 17.9 times higher risk when using 75 ISQ as threshold value. Our data corroborate with the outcomes of other studies showing an increased risk for implant failure with decreased stability [10,15,16]. For instance, in a retrospective study on 300 implants of which 20 were lost after three years, Turkyilmaz and McGlumphy found a significant difference between failed and successful implants when comparing the primary stabilities (46.5 ± 4 ISQ vs 67.1 ± 7 ISQ). Rodrigo et al. evaluated 37 failures of 542 implants after three years of follow-up and found no correlation between primary stability and implant failure but a significant association when measurements were made after a mean period of 2.8 months after placement. is is in line with Atieh et al. who concluded that implant stability measurements after 8 weeks showed a better accuracy in predicting implants that were at risk of failure than those taken at the time of implant placement [39]. e indication for implant treatment influenced the 5year implant survival rate in the present study, i.e., more failures were observed in full jaw than in partial jaw cases. e majority of failures, i.e., seven of eight failures in the full jaw rehabilitations were seen in the maxilla. Both these observations are in line with previous experiences from dental implants [1] and are probably related to morphological differences between the jaws [40]. e RFA analysis showed lower stability for maxillary than for mandibular implants as also observed by other authors [7,27]. Moreover, the ISQ values were lower in full than in partial and single cases. Since the main determinant of RFA is bone density [6,9], these results are in part explained by the fact that soft bone densities are more commonly found in the maxilla than in the mandible. Moreover, it may also be so that implants placed for single or partial prostheses are protected against overload by the remaining dentition, while full jaw constructions have to take the full load during healing and loading with an increased risk for implant failure.
In conclusion, a 5-year CSR of 97.3% was found in the present patient group. More failures were observed in full than in partial rehabilitations. Age, gender, jaw, position, bone quantity, bone quality, implant diameter, and implant Implant failure rate (%) Figure 3: Failure rates for implants with values < or ≥60, 65, 70, and 75 ISQ at placement and loading. Note. ere is a consistent correlation between failure and stability for implants above the threshold levels at placement and loading (see Table 10 for statistical analyses).
length had no influence on implant survival. Gender, jaw, position, bone quality, and implant diameter had an influence on implant stability at placement. Jaw, bone quality, and implant diameter had an influence on stability after 3-4 months of healing. In general, the present study also showed lower survival rates for implants with low stability than for implants with higher stability at the different threshold levels. ere was a significantly higher risk for failure for implants with an ISQ value below 70 and 75 than for implants with higher stability at placement. Moreover, the risk for failure increased further if the ISQ value was still below 70 and 75 after 3-4 months of healing. Since implants with high stability seem to be more successful, it can be speculated that measures to improve implant stability may improve the clinical outcome. However, clinical studies are needed to evaluate the impact of such measures, such as adapted drilling protocols, the use of wider and tapered implants, and prolonged healing periods, on implant stability and implant survival.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Disclosure
Dr. Herman Sahlin is employed by Neoss Ltd., Harrogate, UK, and performed the statistical analyses of the data in the publication.

Conflicts of Interest
e authors declare that they have no conflicts of interest.