High survival rates of dental implants have been found in many long-term clinical trials,1,2 which has led to the current focus of clinical research being on identifying more-specific outcomes, such as periimplant marginal bone.3 Continuous periimplant marginal bone loss can threaten the longevity of the implant-supported prosthesis.4 In particular, a substantial amount of bone loss occurs after the first year of function.5 This early periimplant marginal bone loss is not only an important criterion for success of dental implant but also a crucial factor for the subsequent progression of marginal bone.6 Thus, minimal or no marginal bone loss after connecting the implant and abutment is a critical indicator for the long-term success of implant.7
Many plausible hypotheses have been proposed about early periimplant marginal bone loss, such as surgical trauma,8 biological width,9,10 occlusal overload,11 and microgaps.12 However, there is little evidence for the apparent cause of early bone loss, and few studies have attempted to identify risk indicators while focusing on early bone loss. Prospective longitudinal studies and multivariate statistical analysis are required for identifying the true risk factors.13
Several studies recent have focused on identifying specific host characteristics for periimplant marginal bone loss in addition to those clinical and physiologic factors. Albrektsson et al14 stated that the initial marginal bone loss of dental implants represents an imbalanced host response to the dental implant rather than a disease. Other authors reported that certain host characteristics such as genetic polymorphism15,16 or activities of molecules in the crevicular fluid17 are correlated with periimplant marginal bone loss. Several researchers have recently proposed that human alveolar bone–derived mesenchymal stromal cells (hABCs) are a key factor for the homeostasis of alveolar bone and regenerative strategy due to their differentiation potential in bone metabolism.18,19 We therefore hypothesized that specific genes are involved in bone remodeling, the wound-healing capacity, and early periimplant marginal bone loss. To the best of our knowledge, this is the first clinical study to evaluate the quantitative association between marginal bone loss and the mRNA expression of hABCs by using real-time reverse-transcription polymerase chain reaction (RT-PCR) analysis.
The aim of the prospective cohort pilot study was to identify clinical risk factors relevant to the host and the implant-supported prosthesis and cellular responses of in situ hABCs involved in early periimplant marginal bone loss.
Materials and Methods
Study Design and Participants
In total, 37 patients requiring implant-supported rehabilitation were recruited at the Department of Periodontology, Yonsei University Dental Hospital, between December 2013 and November 2014. All patients who qualify for enrollment were participated in the study. The institutional review board of Yonsei University reviewed and approved the clinical research protocol before the clinical trial begins (Yonsei IRB No. n2-2013-0037). All patients were led to write the informed consent before being enrolled in the study.
The following inclusion criterions were applied: (1) adult male or female (≥18 years) in good systemic health, (2) completely or partially edentulous patients needing dental implantation, (3) patients with pristine periodontal health or well-maintained clinical periodontal health, (4) patients having sufficient alveolar bone volume that could receive dental implants, and (5) patients having stable occlusal relationship. The exclusion criteria were (1) uncontrolled systematic diseases (eg, heart disease or metabolic disease), (2) medications known to interfere with bone metabolism, (3) pregnancy, (4) current smoker or history of smoking within the past 10 years, or (5) extraction sockets with a previous bone grafting procedure. The study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for prospective cohort studies (Supplementary File 1, Supplemental Digital Content 1, http://links.lww.com/ID/A8).20
Implant Placement and Prosthetic Treatment
All patients were treated by 8 skilled surgeons working at the Dept. of Periodontology, Yonsei University Dental Hospital, using the same implant system (Luna; Shinhung, Seoul, Korea). The implant used had a surface with sand-blasted/acid-etched, double thread and internal conical seal design; the diameter of abutments is smaller than that of implant platform (platform switching). All the implants were installed in accordance with the manufacturer's instructions based on appropriate radiographic planning before the surgery. The surgery involved low-speed drilling (50–200 rpm) to obtain bone particles as outlined previously.21 The bone particles were harvested from a single randomly selected site on patients with multiple operation sites (1 sample from each patient). A guided bone regeneration procedure was performed using autogenous bone or xenografts, with resorbable collagen membrane if necessary.
The patients were referred to prosthetic specialists working at the Dept. of Prosthodontic, Yonsei University Dental Hospital, to complete the final prosthesis 3 months or 4 months after implant surgery. A typical maintenance visit for patients was scheduled every 3 months to evaluate any changes in their oral condition, and it lasted for at least 1 hour at every visit. All patients were given the rigorous oral hygiene instruction for cleaning the periimplant soft tissue. Tooth and implant sites that showed easy bleeding on probing were carefully cleaned by using rubber cups in conjunction with antiseptic oral rinses.
The host-related variables were listed as follows: (1) sex (male or female), (2) age (≤60 or >60 years), (3) involved jaw (mandible or maxilla), (4) implant location (posterior [premolar and molar] or anterior [incisor and canine]), (5) bone quality determined by the surgeon who performed the surgery registered in the clinical trial (classification according to Lekholm and Zarb as D2, D3, or D4; to standardize subjectivity of clinical bone quality judgment, all surgeons participated in the calibration meeting before the clinical trial), and (6) relative mRNA expression levels of genes associated with bone remodeling and tissue healing.
The implant- and prosthesis-related variables were (1) implant diameter (narrow [3.5 mm], regular [4.0 or 4.5 mm], or wide [5.0 mm]), (2) implant length (short [7 or 8.5 mm] or long [10 or 11.5 mm]), (3) insertion torque (≤20 or >20 N), (4) initial implant stability quotient (ISQ) (<70 or ≥70), (5) depth of implant placement (subcrestal [with the mesial or distal implant shoulder placed at least 0.5–1 mm]) or equicrestal (with the mesial and distal implant shoulder placed within 0.5 mm below the alveolar crest), (6) presence of Guided Bone Regeneration (GBR) procedure, and (7) type of suprastructure (single crown, fixed partial denture, or overdenture).
RNA Extraction and Quantitative RT-PCR
The isolation procedures performed in the laboratory were as follows: Bone chips obtained by low-speed drilling were put in a 50-mL tube and placed in 20-mL medium comprising a-MEM, 15% Fetal Bovine Serum (FBS) 2-mM L-glutamine, 100-U/mL penicillin and 100-μg/mL streptomycin (Gibco, Invitrogen, Massachusetts, USA), and 100-μM L-ascorbic acid-2-phosphate (Sigma-Aldrich, St. Louis, MO). A sequential digestion method with 2-mg/mL collagenase (Wako Pure Chemicals, Osaka, Japan) and 1-mg/mL Dispase (Gibco, Invitrogen, Massachusetts, USA) was used for the immediate isolation of the hABCs. They were seeded into culture dishes (75T) (Thermo Scientific, Gibco, Invitrogen, Massachusetts, USA), incubated for 3 days at 37°C to allow for cell attachment. To remove any nonadherent cells, the dishes were washed twice with phosphate-buffered saline.
The stem-cell–like characteristics of hABCs at passage 2 or 3 were investigated according to Park et al.21 The primer set for each differentiation and the tissue-healing markers were designed using the software program (Primer software version 3.0, Applied Biosystems, MA). Each cell was harvested using Trizol (15596018; Gibco, Invitrogen, Massachusetts, USA), and total cell RNA was prepared. cDNA synthesis from the isolated total RNA was performed using the High Capacity RNA-to-cDNA kit (4387406; Applied Biosystems). The subsequent PCR was conducted using the PCR system (StepOne real-time PCR system; Applied Biosystems) and the PCR kit (Power SYBR Green PCR Master kit; Applied Biosystems) under the following conditions: an initial polymerase activation step at 95°C for 10 minutes, followed by 40 amplification cycles comprising denaturation at 95°C for 15 seconds and annealing and extension at 60°C for 60 seconds. The relative levels of mRNA expression were quantified by comparison with an internal standard (glyceraldehyde 3-phosphate dehydrogenase). With the same total DNA, each analysis was performed in triplicate. The bone-remodeling– and tissue-healing–associated genes in alveolar bone were grouped into the following categories (Table 1): (1) osteogenesis, (2) osteoclastogenesis, (3) adipogenesis, and (4) tissue healing.
Radiographic Evaluation of Periimplant Marginal Bone Loss
To obtain standardized periapical radiographs, a long-cone parallel technique and a digital imaging software system immediately after implant placement and at 3-month and 1-year follow-up visits (Fig. 1). Care was taken with XCP film holding device (Dentsply Rinn; Elgin, IL) to locate the x-ray film parallel to the long axis of the implant and clearly show the threads on both sides of the fixture. Two researchers, who were not involved in the surgery, independently measured the radiographs (D.-J.K. and J.-K.C.). To avoid introducing any bias into the observations, all the measurements were made on images at the same magnification using image-processing software (Photoshop 12.0, Adobe Systems, CA). For each implant, the loss of marginal bone height was determined as the distance between the first visible bone-to-implant contact and the fixture shoulder. If marginal bone was seen above the fixture shoulder, the measurement was recorded as 0 to avoid introducing bias into the results. To adjust for any magnification error, the measured distance was calibrated based on the ratio of the measured implant length and the actual implant length. All measurements were made at the mesial and distal surfaces of each implant. The intraclass coefficient was calculated to evaluate the consistency between the 2 examiners; it was found to be 0.95 (P = 0.002). The amount of periimplant marginal bone loss over 3 months and 1 year was determined.
The incidence of periimplant marginal bone loss greater than 1 mm was normalized to the total number of implants placed. Differences in the incidence of periimplant marginal bone loss greater than 1 mm between the categories of each risk factor (age, sex, involved jaw, implant location, bone quality, implant diameter, implant length, insertion torque, initial ISQ, implant placement depth, GBR procedure, and type of suprastructure) were analyzed by the chi-square test or Fisher exact test. Statistically relevant risk factors were identified by performing multiple logistic regression analyses. Variables with a probability value of 0.2 or less in the univariate analysis were entered into the model using backward selection with the significance level for removal set at 0.1.22 To facilitate precise interpretations, the odds ratio (OR), the SEM, and 95% confidence interval (CI) for each risk factor are presented.
The statistical analysis of mRNA expression of bone-remodeling– and tissue-healing–associated genes was performed at the patient level. The mRNA data of each group are shown as median and interquartile values due to their highly non-Gaussian distribution. Differences in the expression level between with and without marginal bone loss groups were tested using the nonparametric Mann-Whitney U test for unpaired data.
All statistical analyses were performed using the statistics software (IBM SPSS Statistics [version 21], IBM Corporation, Chicago, IL) with the significance level set to α = 0.05.
The 1-year follow-up was completed in 31 of the 37 (86.5%) patients with 98 implants. The 6 dropouts were due to personal reasons, withdrawal of consent before surgery (n = 5) and not attending the clinic on a scheduled visit (n = 1). The included-subject pool comprised 17 men (54.8%) and 14 women (45.2%). The age of the patients ranged from 22 to 74 years, with a mean of 53 years. The follow-up time was 12.4 ± 0.7 months (mean ± SD).
Implant Survival and Overall Periimplant Marginal Bone Loss
One hundred dental implants were eligible for examination after 3 months and 98 implants after 1 year. None of the implants failed among patients evaluated over the entire study period, representing a 100% survival rate. All the prostheses (53 single crowns, 13 fixed partial denture, and 5 overdenture) survived and remained stable without any complications.
The mean periimplant marginal bone loss (n = 98) was 0.05 ± 0.13 mm at the 3-month follow-up and 0.30 ± 0.60 mm at the 1-year follow-up. Overall, 9 implants (9.2%) showed more than 1.0 mm of the mean periimplant marginal bone loss (1.91 ± 0.74 mm).
Clinical Factors Affecting Periimplant Marginal Bone Loss
Table 2 presents the results of the univariate analysis of the incidence of early periimplant marginal bone loss (>1 mm) of predictor variables. The rate of periimplant marginal bone loss was highest for implants supporting overdentures (50.0%).
The rate of early periimplant marginal bone loss differed significantly with implant location (P = 0.005), bone quality (P = 0.048), and type of suprastructure (P < 0.001), but not with age, sex, involved jaw, implant diameter, implant length, insertion torque, initial ISQ, implant placement depth, or GBR procedure.
From the backward logistic regression analysis of the involved jaw, implant location, bone quality, implant length, and type of suprastructure, the ORs for early periimplant marginal bone loss were significantly higher for overdenture implants (OR = 116.90, SEM = 1.20, 95% CI: 11.09–1232.85, P = 0.001) and those placed in the maxilla (OR = 12.83, SEM = 1.26, 95% CI: 1.08–152.26, P = 0.043). The other variables such as implant location, bone quality, and implant length were excluded from the final model.
After evaluating the amount of periimplant marginal bone loss relative to both the involved jaw and type of suprastructure variables (Fig. 2), the early periimplant marginal bone loss was highest in the maxillary overdenture group (1.72 ± 0.84 mm) and lowest in the mandibular single-crown/fixed-partial-denture group (0.10 ± 0.18 mm). Statistically significant differences were present between (1) the mandibular overdenture group and the single-crown/fixed-partial-denture groups (both in the mandible and maxilla [0.14 ± 0.29 mm]) and (2) the maxillary overdenture group and the mandibular overdenture group (0.49 ± 0.33 mm).
Association Between Early Periimplant Marginal Bone Loss and Relative mRNA Expression of Genes
Among the 31 samples, quantitative RT-PCR was not performed in 2 bone samples due to the failure of RNA extraction and cDNA synthesis; thus, 29 eligible samples were analyzed.
The results of multivariable analysis showed that type of suprastructure (overdenture) was the strongest prognostic clinical factor. To adjust for the confounding effects of environmental factors such as type of suprastructure, the difference in mRNA expression was analyzed in subgroups divided according to the type of suprastructure. Intragroup comparisons in the overdenture subgroup were not performed due to smallness of the sample (n = 2).
Table 3 and Figure 3 indicate that there was an association between early periimplant marginal bone loss and the mRNA expression of bone-remodeling– and tissue-healing–associated genes for single crowns and fixed dental prosthesis (n = 27). Intragroup comparisons were based on the presence or absence of marginal bone loss including minor bone change. There were statistically significant differences in the runt-related transcription factor-2 (Runx-2) (P = 0.041) and bone morphogenetic protein-2 (BMP-2) (P = 0.032) osteogenesis-related genes, and in the peroxisome proliferator-activated receptor gamma-2 (PPARγ-2) (P = 0.025) adipogenesis-related gene and the receptor activator of nuclear factor κ ligand (RANKL)/osteoprotegerin (OPG) ratio (P = 0.041) between the groups with and without marginal bone loss. No significant differences were found in any other bone-remodeling– or tissue-healing–associated genes.
This study focused on identifying clinical factors and cellular responses of in situ hABCs for early periimplant marginal bone loss. To remove the influence of brand-specific characteristics on marginal bone level, we used a uniform implant system having an internal conical connection and platform switching design.
The survival rate was perfect during the first postimplantation year in this study, with no implant or prosthesis complications. Most of the implants (90.8%) showed minimal or no bone loss. These results are consistent with those of previous short-term (6 months after loading or 1 year after insertion) studies using the same implant–abutment connection system.23–25
This study represents the first attempt to determine the quantitative association between the mRNA expression of hABCs and early periimplant marginal bone loss. Runx-2 is essential for osteoblastic precursor cell differentiation, and the null mutation of Runx-2 affects osteoblast differentiation in vivo.26 BMP-2 has also been found to play an important role in osteoblastogenic differentiation, belonging to the transforming growth factor-β superfamily.27 The decreased mRNA expression level of these 2 osteogenesis-related genes in the presence of marginal bone loss may be related to a low potential for bone formation capacity, and it seemed to affect the early marginal bone loss. PPARγ-2 is the key factor driving the adipogenic differentiation of mesenchymal stromal cells and was also decreased in the presence of marginal bone loss.
This similar pattern of mRNA expression in Runx-2, BMP-2, and PPARγ-2 is particularly interesting because it is generally accepted that osteoblastic differentiation and adipogenic differentiation exhibit a reciprocal relationship in normal bone metabolism.28 This can be explained by a dual role of BMP-2 in mesenchymal stromal cell differentiation, and it means that the determination of the osteoblast or adipocyte lineage from a mesenchymal stromal cell precursor is regulated by BMPs and their distinct types of receptors.29 Both RANKL and OPG are produced by osteoblasts and play crucial roles in regulating osteoclastogenesis.30 Their mRNA levels were shown to be correlated with altered bone resorption in response to physiologic stimuli.31 This is consistent with this study finding that the RANKL/OPG ratio was higher (with a high potential for osteoclastic activity) in the presence of marginal bone loss. There is no particular association between the other genes involved in tissue healing and early periimplant marginal bone loss. Because there was no similar study, we were not able to compare these results with other studies.
The multivariate analyses identified the type of suprastructure and the involved jaw as risk indicators for early periimplant marginal bone loss. It was found that the OR and the mean amount of early periimplant marginal bone loss were significantly higher in implants with an overdenture than in implants with a single crown or fixed partial denture. This is consistent with a previous report of a significant association between greater bone loss and the presence of a removable prosthesis.32 By contrast, the systematic review by Bryant et al33 found similar levels of crestal bone loss in the first postimplantation year for both fixed and removable prostheses. In this review, however, the difference in prosthetic type did not reach statistically significant conclusion. Also, there was a lack of research on maxillary removable design, and so, only a site-specific comparison was possible between mandible prostheses. In terms of prosthesis, the clinical influence of suprastructure on marginal bone loss needs high interest, and a randomized-controlled clinical study with high number of patients should be performed to identify it.
Another factor associated with a higher OR for early periimplant marginal bone loss was implant placement in the maxilla. This is consistent with the clinical findings of Vervaeke et al.34 The mean periimplant marginal bone loss of the maxilla overdenture group was significantly higher than that of the mandible overdenture group. In detail, most of the implants with early bone loss were restored with maxillary implant overdentures. It is generally accepted that the maxillary implant overdenture is often correlated with reduced bone quantity/quality and vulnerable on biomechanical forces.35,36 Destructive cantilever forces can be present in the anterior and premolar regions when dental implants collide with natural teeth.37
Some limitations of this study need to be considered when interpreting the results. Its prospective cohort pilot design meant that relatively few patients were treated. Although significant differences were found in the Mann-Whitney U test, the statistical power is lower for nonparametric tests than for parametric tests. Further clinical studies involving larger samples are necessary to properly assess the significance of the results of this study. Another limitation of this study was the use of standardized digital periapical radiographs to evaluate periimplant marginal bone loss. The disadvantage of using periapical radiographs is that these are only 2-dimensional images (in the mesial and distal directions) that might not accurately represent the morphology of the defect site. However, periapical radiography is a widely accepted method used for assessing the interproximal crestal bone changes of dental implants and evaluating the periimplant health.4
Within the limitations of this prospective single-cohort study, the present data demonstrate that some genes involved in bone remodeling (runt-related transcription factor-2 [Runx-2]), bone morphogenetic protein-2 (BMP-2), and peroxisome proliferator–activated receptor gamma-2 (PPARγ-2) and RANKL/OPG ratio RANKL/OPG are correlated with early periimplant marginal bone loss. The suprastructure supporting the implant and the involved jaw were found to be significant clinical risk factors for early periimplant marginal bone loss.
The authors claim to have no financial interest, either directly or indirectly, in the products or information listed in the article.
The institutional review board of Yonsei University reviewed and approved the clinical research protocol before the clinical trial begins (Yonsei IRB No. n2-2013-0037). All patients were led to write informed consent before being enrolled in the study.
Roles/Contributions by Authors
D.-J. Kim: formal analysis and writing original draft. S.-K. Kim: data curation and methodology. J.-K. Cha: formal analysis and reviewing the draft. J.-S. Lee: investigation, reviewing, and editing the draft. C.-S. Kim: research design, investigation, reviewing, and editing the draft.
1. Berglundh T, Persson L, Klinge B. A systematic review of the incidence of biological and technical complications in implant dentistry reported in prospective longitudinal studies of at least 5 years. J Clin Periodontol. 2002;29(suppl 3):197–212; discussion 232–193.
2. Laurell L, Lundgren D. Marginal bone level changes at dental implants after 5 years in function: A meta-analysis. Clin Implant Dent Relat Res. 2011;13:19–28.
3. Tatarakis N, Bashutski J, Wang HL, et al. Early implant bone loss: Preventable or inevitable? Implant Dent. 2012;21:379–386.
4. De Bruyn H, Vandeweghe S, Ruyffelaert C, et al. Radiographic evaluation of modern oral implants with emphasis on crestal bone level and relevance to peri-implant health. Periodontol 2000. 2013;62:256–270.
5. Oh TJ, Yoon J, Misch CE, et al. The causes of early implant bone loss: Myth or science? J Periodontol. 2002;73:322–333.
6. Galindo-Moreno P, Leon-Cano A, Ortega-Oller I, et al. Marginal bone loss as success criterion in implant dentistry: Beyond 2 mm. Clin Oral Implants Res. 2015;26:e28–e34.
7. Hartman GA, Cochran DL. Initial implant position determines the magnitude of crestal bone remodeling. J Periodontol. 2004;75:572–577.
8. Esposito M, Hirsch JM, Lekholm U, et al. Biological factors contributing to failures of osseointegrated oral implants. (I). Success criteria and epidemiology. Eur J Oral Sci. 1998;106:527–551.
9. Wallace SS. Significance of the 'biologic width' with respect to root-form implants. Dent Implantol Update. 1994;5:25–29.
10. Abrahamsson I, Berglundh T, Lindhe J. The mucosal barrier following abutment dis/reconnection. An experimental study in dogs. J Clin Periodontol. 1997;24:568–572.
11. Misch CE, Suzuki JB, Misch-Dietsh FM, et al. A positive correlation between occlusal trauma and peri-implant bone loss: Literature support. Implant Dent. 2005;14:108–116.
12. Hermann JS, Schoolfield JD, Schenk RK, et al. Influence of the size of the microgap on crestal bone changes around titanium implants. A histometric evaluation of unloaded non-submerged implants in the canine mandible. J Periodontol. 2001;72:1372–1383.
13. Koldsland OC, Scheie AA, Aass AM. The association between selected risk indicators and severity of peri-implantitis using mixed model analyses. J Clin Periodontol. 2011;38:285–292.
14. Albrektsson T, Dahlin C, Jemt T, et al. Is marginal bone loss around oral implants the result of a provoked foreign body reaction? Clin Implant Dent Relat Res. 2014;16:155–165.
15. Dereka X, Mardas N, Chin S, et al. A systematic review on the association between genetic predisposition and dental implant biological complications. Clin Oral Implants Res. 2012;23:775–788.
16. Hamdy AA, Ebrahem MA. The effect of interleukin-1 allele 2 genotype (IL-1a(-889) and IL-1b(+3954)) on the individual's susceptibility to peri-implantitis: Case-control study. J Oral Implantol. 2011;37:325–334.
17. Slotte C, Lennerås M, Göthberg C, et al. Gene expression of inflammation and bone healing in peri-implant crevicular fluid after placement and loading of dental implants. A kinetic clinical pilot study using quantitative real-time PCR. Clin Implant Dent Relat Res. 2012;14:723–736.
18. Matsubara T, Suardita K, Ishii M, et al. Alveolar bone marrow as a cell source for regenerative medicine: Differences between alveolar and iliac bone marrow stromal cells. J Bone Miner Res. 2005;20:399–409.
19. Nishimura M, Takase K, Suehiro F, et al. Candidates cell sources to regenerate alveolar bone from oral tissue. Int J Dent. 2012;2012:857192.
20. von Elm E, Altman DG, Egger M, et al. The strengthening the reporting of observational studies in. epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet. 2007;370:1453–1457.
21. Park JC, Kim JC, Kim YT, et al. Acquisition of human alveolar bone-derived stromal cells using minimally irrigated implant osteotomy: In vitro and in vivo evaluations. J Clin Periodontol. 2012;39:495–505.
22. Bursac Z, Gauss CH, Williams DK, et al. Purposeful selection of variables in logistic regression
. Source Code Biol Med. 2008;3:17.
23. Prosper L, Redaelli S, Pasi M, et al. A randomized prospective multicenter trial evaluating the platform-switching technique for the prevention of postrestorative crestal bone loss. Int J Oral Maxillofac Implants. 2009;24:299–308.
24. Fernandez-Formoso N, Rilo B, Mora MJ, et al. Radiographic evaluation of marginal bone maintenance around tissue level implant and bone level implant: A randomised controlled trial. A 1-year follow-up. J Oral Rehabil. 2012;39:830–837.
25. Moergel M, Rocha S, Messias A, et al. Radiographic evaluation of conical tapered platform-switched implants in the posterior mandible: 1-year results of a two-center prospective study. Clin Oral Implants Res. 2016;27:686–693.
26. Komori T, Yagi H, Nomura S, et al. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell. 1997;89:755–764.
27. Marie PJ, Debiais F, Haÿ E. Regulation of human cranial osteoblast phenotype by FGF-2, FGFR-2 and BMP-2 signaling. Histol Histopathol. 2002;17:877–885.
28. Gimble JM, Robinson CE, Wu X, et al. The function of adipocytes in the bone marrow stroma: An update. Bone. 1996;19:421–428.
29. Wang EA. Bone morphogenetic proteins (BMPs): Therapeutic potential in healing bony defects. Trends Biotechnol. 1993;11:379–383.
30. Krane SM. Identifying genes that regulate bone remodeling as potential therapeutic targets. J Exp Med. 2005;201:841–843.
31. Thomas GP, Baker SU, Eisman JA, et al. Changing RANKL/OPG mRNA expression
in differentiating murine primary osteoblasts. J Endocrinol. 2001;170:451–460.
32. Tandlich M, Ekstein J, Reisman P, et al. Removable prostheses may enhance marginal bone loss around dental implants: A long-term retrospective analysis. J Periodontol. 2007;78:2253–2259.
33. Bryant SR, MacDonald-Jankowski D, Kim K. Does the type of implant prosthesis affect outcomes for the completely edentulous arch? Int J Oral Maxillofac Implants. 2007;22(suppl):117–139.
34. Vervaeke S, Collaert B, Cosyn J, et al. A multifactorial analysis to identify predictors of implant failure and peri-implant bone loss. Clin Implant Dent Relat Res. 2015;17(suppl 1):e298–307.
35. Chan MF, Narhi TO, de Baat C, et al. Treatment of the atrophic edentulous maxilla with implant-supported overdentures: A review of the literature. Int J Prosthodont. 1998;11:7–15.
36. Rodriguez AM, Orenstein IH, Morris HF, et al. Survival of various implant-supported prosthesis designs following 36 months of clinical function. Ann Periodontol. 2000;5:101–108.
37. Rangert B, Krogh PH, Langer B, et al. Bending overload and implant fracture: A retrospective clinical analysis. Int J Oral Maxillofac Implants. 1995;10:326–334.