Associations among Primary Stability, Histomorphometric Findings, and Bone Density: A Prospective Randomized Study after Alveolar Ridge Preservation with a Collagen Cone

Background: The objective of this investigation was to examine whether determination of bone density (BD) with a cone beam computed tomography (CBCT) scan could help predict the primary stability (PS) of the implants and to investigate whether associations between the histomorphometric findings and the CBCT scan could be observed. Materials and methods: In this randomized clinical study, the efficacy of alveolar ridge preservation (ARP) with a combination of a collagen cone and a collagen membrane procedure after tooth extraction was investigated. CBCT scans were obtained after a healing period of 8 (±1) weeks. Subsequently, the CBCT scans were evaluated in terms of BD at different heights of the former socket. Eleven (±1) weeks after tooth extraction, implant placement was performed and PS was measured with resonance frequency analysis. Potential associations among the radiologically measured BD, the histomorphometric results, and the PS were analyzed. Results: No direct association was observed between the radiologically determined BD and the histomorphometric findings. No significant associations could be found between the BD and the PS. Conclusion: No significant associations were observed among the BD determined by the CBCT, the histomorphometric findings, and the PS.


Introduction
Primary stability of implants is considered a prognostic factor for osseointegration. Concepts of early implant loading also require sufficiently high primary stability. Therefore, information about bone quality obtained beforehand using measures such as cone beam computed tomography (CBCT) could be helpful.
For the evaluation of bone density in Hounsfield units (HUs), computed tomography (CT) scans have to be obtained, as CBCT scans with reduced radiation exposure are not suitable [1]. CBCT scans have limitations while analyzing the quantitative data of bone density measurements [2]. Furthermore, the measurement of bone density depends significantly on the CBCT device, and on the scanning

Participants
In total, 88 patients with at least one tooth in the upper jaw indicated for extraction and replacement by an implant-supported prosthetic restoration participated in the study. Patients were consecutively divided into two groups (the test group and the control group, 44 patients each). Participants in the pilot study and the main study were divided into three subgroups. Twenty patients underwent histological examination [15]. Eight patients were included in the test of an implant insertion tool. Sixty patients were included in the main study for implant placement with a modified implant drill system and follow-up examination ( Figure 1). The division into subgroups was necessary because the drilling protocol of the manufacturer was violated by sampling for histological examinations. Furthermore, two different versions of implants were used (Camlog Guide 1.0 vs. 2.0).

Materials and Methods
The study design corresponded to the design described in the previously published article by Schnutenhaus et al. [24]. The present trial was performed as q parallel-group randomized human clinical trial in accordance with the Declaration of Helsinki. After approval from the Ethical Committee of Ulm University (No. 337/12, 13 February 2013) the study was registered in the German Clinical Trials Register as DRKS00004769 (International Clinical Trials Registry Platform of the World Health Organization).
The study participants were informed about the study before their participation orally as well as in writing and a written informed consent was obtained from all participants.

Participants
In total, 88 patients with at least one tooth in the upper jaw indicated for extraction and replacement by an implant-supported prosthetic restoration participated in the study. Patients were consecutively divided into two groups (the test group and the control group, 44 patients each). Participants in the pilot study and the main study were divided into three subgroups. Twenty patients underwent histological examination [15]. Eight patients were included in the test of an implant insertion tool. Sixty patients were included in the main study for implant placement with a modified implant drill system and follow-up examination ( Figure 1). The division into subgroups was necessary because the drilling protocol of the manufacturer was violated by sampling for histological examinations. Furthermore, two different versions of implants were used (Camlog Guide 1.0 vs. 2.0). Forty-four patients received a combination of a collagen cone and a collagen membrane (Parasorb Sombrero ® ; Resorba, Nürnberg, Germany), which was inserted into the alveoli after tooth extraction (test group). For patients in the control group, wound healing was allowed to proceed without further intervention. Inclusion and exclusion criteria are listed in detail in the previously published article by Schnutenhaus et al. [24]. The clinical therapy steps were carried out between February 2013 and March 2015.

Treatment Protocol
The monocentric study was conducted at the private practice of the first author (SIS), who exclusively performed all interventions and follow-ups. All participating patients were recruited from the same private practice in Hilzingen, Germany.
Interventions on the day of tooth extraction included local anesthesia with articaine (Ultracain DS 1:200,000; Sanofi Aventis, Frankfurt, Germany), atraumatic extraction of the teeth after complete mobilization, curettage of the extraction socket, and no further measures for the patients in the control group. Patients in the test group received a combination of a collagen cone and a collagen membrane Forty-four patients received a combination of a collagen cone and a collagen membrane (Parasorb Sombrero ® ; Resorba, Nürnberg, Germany), which was inserted into the alveoli after tooth extraction (test group). For patients in the control group, wound healing was allowed to proceed without further intervention. Inclusion and exclusion criteria are listed in detail in the previously published article by Schnutenhaus et al. [24]. The clinical therapy steps were carried out between February 2013 and March 2015.

Treatment Protocol
The monocentric study was conducted at the private practice of the first author (SIS), who exclusively performed all interventions and follow-ups. All participating patients were recruited from the same private practice in Hilzingen, Germany.
Interventions on the day of tooth extraction included local anesthesia with articaine (Ultracain DS 1:200,000; Sanofi Aventis, Frankfurt, Germany), atraumatic extraction of the teeth after complete mobilization, curettage of the extraction socket, and no further measures for the patients in the control group. Patients in the test group received a combination of a collagen cone and a collagen membrane according to the manufacturer's instructions. All patients were given post-extraction instructions for the 24 h after tooth removal. One week later, visual inspection of all wounds was performed.
A nonsteroidal anti-inflammatory drug (600 mg ibuprofen) was prescribed for pain reduction, to be taken as needed. No systemic prophylactic antibiotic was performed.

Radiological Measurement of Bone Density
Overall, 88 patients were enrolled in the study. CBCT scans were used to analyze bone density in the region of the extraction socket after a healing period of 8 (±1) weeks (T1).
CBCT scans for the evaluation of bone density were loaded into the software OsiriX Lite Version 9.0 (Pixmeo SARL, Bernex, Switzerland). The dentists analyzing the scans were calibrated prior to the evaluation. Each measurement was performed with the same computer at a single workspace. Measurements were performed at three different heights of the extraction sockets. The first measurement was at the junction of the coronal third and the lower two thirds, the second was in the middle, and the third measurement was at the junction of the apical third and the upper two thirds of the alveoli. A fourth measurement of the local bone was further made apical and palatal to the alveoli ( Figure 2). For each measurement, a region of interest with an area of 2 mm 2 was placed in the center of the alveoli ( Figure 3).
according to the manufacturer's instructions. All patients were given post-extraction instructions for the 24 h after tooth removal. One week later, visual inspection of all wounds was performed.
A nonsteroidal anti-inflammatory drug (600 mg ibuprofen) was prescribed for pain reduction, to be taken as needed. No systemic prophylactic antibiotic was performed.

Radiological Measurement of Bone Density
Overall, 88 patients were enrolled in the study. CBCT scans were used to analyze bone density in the region of the extraction socket after a healing period of 8 (±1) weeks (T1).
CBCT scans for the evaluation of bone density were loaded into the software OsiriX Lite Version 9.0 (Pixmeo SARL, Bernex, Switzerland). The dentists analyzing the scans were calibrated prior to the evaluation. Each measurement was performed with the same computer at a single workspace. Measurements were performed at three different heights of the extraction sockets. The first measurement was at the junction of the coronal third and the lower two thirds, the second was in the middle, and the third measurement was at the junction of the apical third and the upper two thirds of the alveoli. A fourth measurement of the local bone was further made apical and palatal to the alveoli ( Figure 2). For each measurement, a region of interest with an area of 2 mm 2 was placed in the center of the alveoli ( Figure 3).

Primary Stability of the Implants
All implants were inserted using a template-guided procedure according to the study protocol and the manufacturer's instructions. No further modification in the implant position was made, as the accuracy of this procedure was analyzed in a subsequent study. The primary stability was measured immediately after the removal of the surgical template. Insertion posts were removed carefully while stabilizing the implant with a counter ratchet to prevent derotation. The implant stability quotient (ISQ) values were measured with Osstell TM (Integration Diagnostics AB, Göteborg, Sweden), a device that uses RFA technology. Two measurements from two different directions (mesial and buccal) were collected for each implant.

Association between the Primary Implant Stability and the Radiological Bone Density Measurement
A positive association between the primary implant stability and the radiologically measured bone density was expected. The results of the association analysis were expected to serve as a reference for the expected primary stability after CBCT analysis.

Histomorphometric Evaluation
A semiquantitative evaluation was performed based on the qualitative characteristics. The first step was to test whether the features were recognizable. The results were then classified semi-quantitatively. As a result, a yes/no decision was logged.
The samples were evaluated for osteogenesis, remodeling, osteoblast activity, discernible calcification/mineralization, detection of bundle bone in the alveoli, vascularization of the alveoli, and inflammation. Materials and methods are described in detail in the previously published article by Schnutenhaus et al. [15]. The number of cases with 10 test and 10 control samples were based on the specifications of the ISO10993-6 [25].

Randomization
A randomization list was created for the entire study (Institute of Epidemiology and Medical Biometry, University of Ulm, Ulm, Germany). Assignment to the various groups was made in six layers. The data were stratified as follows: • According to sex (2 groups: male and female) • According to the region of the test tooth (3 groups: anterior, premolar, and molar) A responsible person informed the study center in which group of patients was classified according to the randomization list.

Blinding
Blinding of the socket treatment was not possible due to the study design. However, CBCT images taken at T1 were forwarded to the analyst (CE) in blinded and anonymized form. The histological evaluation (WG) was also blinded. Unblinding was performed only after the completion of the analysis, documentation, and statistical analysis.

Statistical Analysis
For the continuous variables, minimum, median, and maximum values were reported. Nominal and ordinal features were described with their absolute and relative frequencies.
Differences between the test and the control groups were investigated using the Wilcoxon rank-sum test. Due to the exploratory nature of this study, all statistical results were hypothetical in nature and should not be interpreted as confirmatory. All statistical tests were carried out at level α = 0.05 (two-sided). No adjustments were made for multiple testing.
The association analysis of the measured ISQ values and the adjusted values of bone density measurements was performed using scatter plots and Spearman's rank correlation coefficient for two continuous variables and the point-biserial correlation coefficient for continuous and binary variables.
Due to a lack of clinical data, sample size could not be estimated in advance. Therefore, this study as performed as an explorative study.

Study Population
This study included 88 patients for the collective analysis. One drop-out was a patient who did not show up for tooth extraction after having been recruited. All patients were treated according to the clinical protocol without any post-operative complications. All included patients completed the trial. Patients were assigned to groups in six strata. Forty-four patients were randomized to be treated with the collagen material and 43 patients served as controls, receiving no further interventions after tooth extraction.
Fifty-two female and 35 male patients participated in this study. The ARP group consisted of 27 female and 17 male patients, while the control group consisted of 25 female and 18 male patients. The mean patient age was 49.6 years (range: 21.7-77.3 years). In the histomorphometric study 10 female patients and 10 male patients were included. The mean age of this patients was 46.6 (21.9-71.4) years. Randomized distribution of the teeth according to their position in the jaw is shown in Table 1. Table 1. Distribution of the teeth of 87 patients according to their positions in the jaw.

Region. ARP Group Control Group Total
Anterior teeth  17  17  34  Premolars  19  21  40  Molars  8  5  13 ARP: alveolar ridge preservation. Table 2 shows the parameters of the radiological bone density as a function of localization of the measuring points in the alveoli for the ARP and the control groups. The mean values were reported for each sectional plane. The values were standardized with the respective value of alveolar bone density measured in the palatal-apical direction. This procedure allowed generation of "Hounsfield-equivalents." The Wilcoxon rank-sum test showed no significant differences between the ARP group and the control group for any of the regions. Table 2. Bone density after tooth extraction in the alveolar ridge preservation group and the control group. Mean values of the respective sectional planes were standardized with the average value of the local alveolar bone density measured in the palatal-apical direction (calculated as "Hounsfield-equivalents"). In contrast, significant differences were found among the bone densities at different sectional planes ( Table 3). The bone density increased from the coronal measuring points to the more apically located measuring points. The difference of 0 analyzed by Wilcoxon signed-rank test was p < 0.01 for all comparisons. There were no significant differences between the ARP group and the control group. Table 3. Differences in bone density among different measurement locations in the alveolus according to the overall group, the ARP-group, and the control group including maxima, median, and minima. The difference of 0 was investigated using Wilcoxon signed-rank test and the group comparison was performed using Wilcoxon rank-sum test.

Association between Radiologically Measured Bone Density and Histomorphometric Results
For the analysis of potential associations between the radiologically measured bone density and the histomorphometric results, scatter plots were applied and Spearman's rank correlation coefficient was calculated for histomorphometric parameters with more than two manifestations. When two manifestations or yes/no records were present, the point-biserial correlation coefficient was calculated (Table 4). No parameter showed significant differences (p < 0.05). Descriptive data revealed pronounced remodeling and higher bone density at the more apically located measuring points. Stronger vascularization or signs of inflammation were associated with reduced bone density at all the sectional levels.

Primary Stability of Implants
Sixty patients from the main study group were included for the evaluation of the primary stability. There were five dropouts including one patient who missed his appointment for tooth extraction after having been recruited (control group), three patients with implant placement failure due to clinically insufficient primary stability (control group), and one patient who had to cancel the treatment due to severe illness (ARP group). All other patients were treated according to the clinical protocol. No postoperative complications were observed.
After stratified randomization according to sex and the region of the studied tooth, 29 patients received ARP measures after tooth extraction. No interventions were performed after tooth extraction in the control group consisting of 26 patients.
Thirty-four female and 21 male patients participated in this study. The ARP group consisted of 19 female and 10 male patients, while the control group consisted of 15 female and 11 male patients. The mean patient age was 51.5 years (range: 24.0-77.0 years). Randomized distribution of the teeth according to their positions in the jaw is shown in Table 5. Table 5. Distribution of the teeth of 55 patients according to their positions in the jaw.
The distribution of implant lengths and the diameters for both groups is presented in Table 6.  Table 7 demonstrates the descriptive data for the primary stability. Primary stability was measured in the buccal and in the mesial direction. No significant differences were found in the mean values between the ARP group and the control group.

Association between the Primary Stability and the Radiologically Measured Bone Density
We also examined whether there was an association between bone density calculated from the CBCT scan and the primary stability of the implants. Table 8 shows the association analysis of these two parameters. Spearman's rank correlation coefficient was calculated. No significant correlations were observed between radiologically measured bone density and the primary stability of the implants. Descriptive data showed higher bone density values and greater primary stability at the more apically located measuring points.

Measurements of Bone Density
No significant difference was observed in the radiologically measured bone density using CBCT scan between the ARP group and the control group. The method applied in the present study is suitable for recording the healing process of the alveolus. As expected, bone density increased significantly from coronal end to apical end. However, no significant difference was found in bone density between the ARP group and the control group at any measurement point of the former socket.
In a previous clinical study, the effect of two ARP measures (a demineralized bovine material and a self-expanding composite) was compared after six months. Results showed significant differences in the radiologically measured bone density in the apical and the middle parts of the alveolus [26]. Implant placement three months after the extraction represents a clinically proven concept [27]. This period was also selected in the present study. During this period, new bone formation was found histologically [15]. However, mineralization that has an impact on the radiologically assessable bone density had not yet occurred. Radiologically measured differences in the density can also be influenced by residual bone grafting material. In the present study, this error was avoided, as the ARP material was completely absorbable. Histological investigations on bovine material have demonstrated that 30.8% of the overall bone volume consisted of the bone substitute nine months after the ARP measure [28].
Determination of bone quality with the help of CBCT scans shows reasonable reliability and validity. Further histomorphological evaluations to analyze the trabecular structures or to calculate the bone volume in proportion to the overall volume can be achieved using micro CT scans [29]. In an animal study, no association could be observed between bone density measured using a CBCT scan and the bone quality in the upper jaw as opposed to the lower jaw [30]. This fact might be another possible limitation of this procedure. Pauwels et al. also suggested that the attempt to correct the gray values of a CBCT scan is not sufficient for bone quality assessment [1].
Effects of ARP measures on bone density can be evaluated with CBCT scans obtained for implant planning if values are normalized by means of a control group [31]. Salimov et al. identified a significant association between bone density measured using a CBCT scan and the primary stability of implants measured using RFA. They suggested that CBCT scans are suitable to predict the implant stability and, therefore, help decide whether immediate or early loading of implants should be performed [22].
CT scans do not appear to have any advantage over CBCT scans in classifying bone quality [32]. The mean gray values obtained by a CBCT scan and HUs in a CT scan show a strong correlation [33]. Contrary to results of the present investigation, the study by Batsami et al. showed that the gray values from a CBCT scan of rabbits can be used to assess bone quality and the amount of new bone formation before implant placement [34]. Gonzáles-Garcia and Monje found a significant correlation between the gray values of a CBCT scan and the bone density in biopsies of the upper jaw. Therefore, they believed that clinical bone quality can be assessed using a pre-operative CBCT scan [35].

Primary Stability of Implants
In the present study, the insertion of a collagen cone with a collagen membrane did not significantly improve the primary stability of implants placed 11 weeks after the extraction. Results of RFA in the present study are similar to previously published values. Lozano-Carrascal et al. found average ISQ values of 67.2 ± 4.42 and 49.2 ± 15.3 for tapered and cylindrical implants, respectively [36]. The macro-design of the implants used in the present investigation was conical. Implants with diameters of 3.3 mm, 3.8 mm, or 4.3 mm and with lengths of 7 mm, 9 mm, 11 mm, and 13 mm were placed. According to Ohta et al., the diameter of conical implants does not influence implant stability [37], whereas the implant length is positively associated with the ISQ values of conical implants [38].
Winter et al. performed a finite element analysis of different factors influencing primary stability. The analysis included implant length and thickness of the cortical bone [39]. A positive impact of the thickness of the crestal and the buccal cortical bone on the primary stability of implants was also observed in clinical studies [40]. In vitro, a positive correlation could be observed between the RFA value and the bone implant contact (BIC) [41]. However, this could not be demonstrated in animal experiments. Therefore, conclusive association between RFA values and the primary BIC has not been proven [42]. In a histological study, a significant positive correlation was observed between the bone volume and RFA values [43]. However, in the present study, no significant correlations were observed between the histomorphometric results and bone density measurements and between bone density measurements and the primary stability of implants.
Prognostic value of the primary stability on the survival of implants does not seem conclusive. In contrast, there is a significant correlation between the primary and the secondary stability [18]. Immediate loading of implants requires high primary stability [44]. Therefore, radiological preoperative assessment of primary stability would be helpful in treatment planning. Further conclusive clinical studies regarding bone quality assessed by CBCT scans need to be carried out. Funding: The work was supported by Oral Reconstruction Foundation (former Camlog Foundation) Grant No. CF 41.305. It was self-funded by the authors and their institutions in its major parts. Study design, data collection and data analysis, as well as the writing of the manuscript was carried out completely independently by the authors. The sponsors had no role in the design, execution, interpretation, or writing of the study.