1.15.1 The implant-bone interface
Many studies have assessed the relationship between bone quality and RFA data. Early studies compared cutting resistance at time of osteotomy preparation and implant placement. Friberg et al (1999) carried out a study on maxillary implants comparing cutting resistance and RFA, coupled with a 20-month follow up of RFA values. The two considerable findings of this study were that firstly there was a relationship between increased cutting resistance, which could indicate high bone density, and increased RFA values. Secondly, after 1 year in function all implants tended to reach similar RFA values, despite initial variation in stability values. Similar findings were demonstrated by Andersson et al (2008) in their study of 102 implants. An inverse relationship between bone-density demonstrated by cutting torque, and changes in implant stability were found over a one-year follow-up. In addition there was a correlation between bone quality and initial implant stability, where implants placed in soft bone with a low primary stability showed a marked increase in stability over 12 months in comparison to more subtle changes in RFA values of implants placed in dense bone with initial high stability. The findings regarding the relationship between bone quality and density and initial RFA values indicate that the stiffness of the implant-bone interface is higher
validated by the Turkyilmaz et al (2009) cadaver study, in which a positive correlation was found between cutting insertion torque and RFA values, in addition to RFA values and Hounsfield Units analyzed by computed tomography scans.
In conjunction with many other authors, Ostman et al (2006) established that RFA values are routinely found to be higher in the mandible than the maxilla, these differences are thought to be caused by relative differences in the density of maxillary and mandibular bone. Other findings from this study of 905 consecutively placed implants showed that stability could also be dependent on gender, implant diameter and the anterior / posterior position of the implant; and that implants of increasing length often showed lower RFA values.
Conversely, a study by Huwiler et al (2007) found there was no correlation between ISQ values and bone density and/or bone trabecular connectivity based on use of micro-CT analysis of osteotomy cores taken with a trephine prior to placement of 23 Straumann™ implants. The authors speculated that ISQ levels do not in fact give an accurate representation of the bone-implant interface at the time of implant placement; but more likely reflect the density and thickness of the cortical layer of the alveolar crest.
Studies have suggested an increased amount and thickness of the cortical bone surrounding an implant at time of placement has a positive effect on RFA values. In a study of 225 implants placed in both the maxilla and mandible and analyzed via computed topography, Myiamoto et al (2005) was able to demonstrate a correlation between available cortical bone and RFA values. Similarly, Nkenke et al (2003) was able to demonstrate a positive correlation between crestal cortical bone height and RFA values in their study on cadavers.
1.15.2 Marginal bone loss
In the initial Meredith et al (1996) study on the use of RFA as a method to test implant stability, it was shown that RFA could be used to differentiate between implants mounted at differing levels in aluminum blocks. This early study suggested that RFA may be suitable for the assessment of marginal bone levels or loss around implants. Sennerby et al (2005) were able to demonstrate that even small changes in marginal bone levels could be detected by changes in RFA. This study evaluated experimentally made marginal bone defects in the dog model and evaluated changes in RFA values with radiographic and histological analysis; results demonstrated that even a minor loss in marginal bone levels resulted in a decrease in electronic RFA values.
In their 5 year study of 52 maxillary implants, Meredith et al (1997) found a positive correlation between bone loss and decreased resonance frequency values by measuring the number of radiographically exposed threads after 5 years and the comparative electronic resonance frequency values. Turkyilmaz et al (2006) found a similar correlation between marginal bone loss and decreasing RFA values in mandibular implants in the first 6 months following implant placement. However no such correlation was seen after the 6 month mark in this 1 year follow-up study; the authors suggested that the increase in stability after 6 months may be due to a compensatory effect of new bone formation and remodeling around the remaining covered implant surface. Conversely, Fischer et al (2008) found no correlation in the relationship between marginal bone loss and RFA values in the first year post implant placement. However, following 3 to 5 years of follow up the same authors found a strong correlation between marginal bone loss and decreasing RFA values.
In 2018 a study was conducted in the beagle dog model assessing if there was a correlation between marginal bone loss caused by ligature induced PI and changes in ISQ values over time. 36 implants were subjected to ligature-induced PI, with marginal bone loss increasing and ISQ levels decreasing with time resulting a strong negative correlation (r=-0.58, p <0.001) at all 4 tested timepoints. A decrease of 1 ISQ point equated to approximately 1mm of
marginal bone loss, despite ISQ values failing to show correlations with any other clinical parameters recorded (Monje et al, 2018).
1.15.3 Orientation of transducer
In an ideal situation an implant should be surrounded by ample bone in all planes, however this is not necessarily always the case. Bone may often be thinner in the bucco-lingual than the mesio-distal dimension and this may result in discrepancies between relative RFA values. Early generation RFA machines often produced differing readings dependent upon the direction the magnetic or electronic force was applied. Veltri et al (2007) discovered in their study that ISQ levels may differ by up to 10 points dependent on whether the transducer was placed in a mesio-distal or bucco-lingual direction.
Some newer generations of RFA machines such as the Osstell Mentor™ take averages of such measurements to ensure continuity in readings taken from different directions, however this suggests caution should be taken when comparing readings from machines of different
generations. Valderamma et al (2007) completed a randomized controlled trial comparing 777 RFA measurements on original electronic RFA (ERFA) machines and 711 measurements with newer magnetic RFA machines (MFRA). They found a mean difference of 8-12 points with higher measurements being found on newer MFRA machines compared to the older ERFA.
This again reiterates caution should be taken when comparing results from different generations of or different forms of resonance frequency analysis.