BRONJ, DENTAL IMPLANTS,
Bisphosphonate-Related Osteonecrosis Associated With Dental Implants: A MicroComputed Tomography Study Luis Junquera, MD, DDS, PhD,* Alejandro Pelaz, MD, DDS, PhD,† Lorena Gallego, MD, DDS, PhD,‡ Luis García-Consuegra, MD, DDS, PhD,§ and Serafín Costilla, MD, DDS, PhDk
isphosphonates (BPs) are nonmetabolic synthetic analogs of pyrophosphate and have a signiﬁcant effect on bone turnover, inhibiting bone resorption by suppressing the activity of osteoclasts,1,2 which results in modiﬁcations of bone microarchitecture. BPs are metabolically inactive so they bind to bone, creating a reservoir that remains even after discontinuation of treatment.3 BPs can be administered orally or intravenously. Intravenous BPs are primarily used in the management of solid tumors, such as breast, prostate, or lung cancer, but are also used to treat lytic lesions in patients with multiple myeloma. These patients often experience sequential skeletal complications, which may include pain, hypercalcemia of malignancy, spinal cord compression, and pathologic fracture.4 Patients with metabolic bone disorders such as osteoporosis and Paget disease of bone similarly have decreased bone strength and increased susceptibility to
*Professor, Department of Maxillofacial Surgery, Central University Hospital, University of Oviedo, Oviedo, Spain. †Department of Maxillofacial Surgery, Cabueñes Hospital, Gijón, University of Oviedo, Spain. ‡Department of Maxillofacial Surgery, Cabueñes Hospital, Gijón, University of Oviedo, Spain. §Department of Maxillofacial Surgery, University of Oviedo, Central University Hospital, Spain. kProfessor, Chief of Department of Radiology, Central University Hospital, University of Oviedo, Oviedo, Spain.
Reprint requests and correspondence to: Luis Junquera, MD, DDS, PhD, Facultad de Ciencias de la Salud, Universidad de Oviedo, Catedrático José Serrano s/n, Oviedo 33009, Spain, Phone: 0034 985 10 35 30, Fax: 0034 985 10 62 85, E-mail: [email protected]
ISSN 1056-6163/14/02303-258 Implant Dentistry Volume 23 Number 3 Copyright © 2014 by Lippincott Williams & Wilkins DOI: 10.1097/ID.0000000000000103
Purpose: The aims of this study were to quantify bone mineral density (BMD) and trabecular architecture of sequestra in patients with bisphosphonate-related osteonecrosis of the jaw (BRONJ) associated with dental implants and to assess if there are differences between trabecular bone in “implant loadingtriggered” and “implant surgerytriggered” BRONJ. Materials and Methods: Bone sequestra of 2 patients diagnosed with BRONJ associated with dental implants were scanned using highresolution microcomputed tomography (microCT). Images were obtained at a voltage of 50 kV and 800 mA, and the specimens were scanned at 180 degrees with a single rotation step of 0.3, 1-mm aluminum ﬁlter, and a pixel size of 12 mm. The morphometric parameters examined were: BMD, ratio of bone volume/tissue volume (BV/TV),
trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular separation (Tb.Sp), connectivity density (Conn.D, mm−3), degree of anisotropy, and the structural model index. Results: BMD and BV/TV were higher in bone sequestration than in healthy bone. Tb.Sp was lower and Tb.N and Tb.Th were higher in the BRONJ group. Conn.D and Tb.N values were signiﬁcantly high in implant surgery-triggered sequestrum but substantially low in sequestra caused by loading as compared with those of the control sample. Conclusions: MicroCT is useful for assessing bone sequestration of BRONJ associated to dental implants. The necrotic bone is similar to that described in conventional BRONJ. (Implant Dent 2014;23:258–263) Key Words: BRONJ, bisphosphonates, micro-computed tomography, bone sequestrum, dental implants
bone pain and fracture. Oral BPs are approved to treat osteoporosis and are frequently used to treat osteopenia as well. In 2009–2010 alone, over 1 million people in the United Kingdom were prescribed BPs, and over 6.5 million prescriptions were issued.5 Bisphosphonate-related osteonecrosis of the jaw (BRONJ) is an uncommon but potentially serious complication of BP therapy. The US Food and Drug Administration received the ﬁrst spontaneous
reports of osteonecrosis in patients with cancer treated with BPs in 2002, and the ﬁrst published reports of BRONJ appeared in the literature in 2003.6,7 Researchers have recently discovered additional drugs that may be associated with a risk for osteonecrosis, including the monoclonal antibodies denosumab and bevacizumab and the multikinase inhibitor sunitinib.8 BRONJ associated with dental implants has been extensively documented
IMPLANT DENTISTRY / VOLUME 23, NUMBER 3 2014 in recent literature.9–13 Several cases have shown that not only the surgical procedure but also subjecting the implant to functional loading can be associated with BRONJ.14 Quantitative bone morphometry is a method used to assess the structural properties of trabecular bone.15,16 The microstructure of trabecular bone consists of various combinations of plates and rods. A 3-dimensional (3D) approach to the analysis of complex trabecular structure is ideally required, as there are limitations in traditional 2-dimensional (2D) histological analysis. The conventional approach to morphologic measurements typically requires a signiﬁcant preparation of the sample, which includes embedding in methylmethacrylate to section it into slices. This technique offers high spatial resolution and image contrast, but, however, it is tedious and time consuming. Particularly inconvenient is the destructive nature of this method, which prevents the sample from being used for other studies, like multiplane analysis or mechanical testing. Microcomputed tomography (microCT) is a new method that enables bone imaging and quantiﬁcation with very high resolution (1 mm), which allows 3D studies of bone microarchitecture and its role in bone mechanical properties. This technique performs a nondestructive quantitative study of bone morphometry and also evaluates other important microstructural features when assessing the mechanical integrity of trabecular bone, such as the incidence and prevalence of microcallus formations. MicroCT has been reported as the preferred method to evaluate trabecular bone microstructure.17,18 This study had 2 primary objectives. One aim was to quantify bone mineral density (BMD) and trabecular architecture of sequestra in patients with BRONJ associated with dental implants. The other aim was to assess if there are differences between trabecular bone in “implant loading-triggered” osteonecrosis and “implant surgery-triggered” BRONJ. The latter is a condition developed within 5 months after surgery, which suggests this procedure is a contributing factor.
Patients and Biopsy Specimens
Four determinations of 2 bone sequestra from 2 patients with conﬁrmed BRONJ associated with dental implants were scanned using microCT. As a control, we used mandibular bone fragments of a healthy 45-year-old man, obtained during a routine dental extraction. Two areas of interest were determined for each of the fragments, one in the basal mandibular bone and the other in the alveolar process located closest to the mouth (Fig. 1). In one of the patients (Figs. 1–3), osteonecrosis (Stage II) was located around an implant subjected to prosthetic loading 5 months before zoledronic acid administration (Zometa; Novartis Pharmaceuticals, East Hanover, NJ). BRONJ appeared 18 months after functional loading. In the other patient, osteonecrosis (Stage III) appeared 5 months after surgery. This patient had been taking alendronate (Fosamax; Merck, Whitehouse Station, NJ) 48 months before implant surgery (Figs. 4–6). Table 1 shows the characteristics of both cases. The study was reviewed and approved by the Ethical Committee of the Central University Hospital of Asturias, Spain. Written consent of each patient was also obtained. Bone Mass and Measurement of 3D Architectural Parameters
All samples were scanned in the Biotery of the University of Oviedo using high-resolution microCT systems (SkyScan 1174; SkyScan, Kontich, Belgium). Images were obtained at a voltage of 50 kV and 800 mA. Samples
Fig. 1. Sagittal reconstruction from case number 1 with the description of the 2 areas of study: (1) alveolar bone and (2) basal bone.
Fig. 2. Case 1. BRONJ surrounding distal implant in left mandibular region (red arrows).
Fig. 3. Case 1. Excised bone sequestration studied including a dental implant in thickness.
were positioned in a specimen holder to keep them perpendicular to the ﬂoor and the x-ray beam, with a maximum length of 50 mm and a maximum width of 30 mm for each of the samples. They were scanned at an angle of 180 degrees with a 0.3 degrees single rotation step, 1-mm aluminum ﬁlter, and 12 mm pixel size. Scanning time for each sample was approximately 3 hours using 9000 ms exposure time and 2-frame average. Flat
Fig. 4. Case 2. Osteolysis affecting the right anterior mandibular area in relation to the dental implant.
BRONJ, DENTAL IMPLANTS,
Fig. 5. Case 2. Mandibular bone defect after removing the dental implant and the necrotic bone.
Fig. 6. Case 2. Extraoral ﬁstula (BRONJ Stage III of Ruggiero et al).
ﬁeld correction was applied before every scan. All images were subjected to 3D reconstruction using NRecon software (SkyScan). Attenuation coefﬁcient, beam hardening, and smoothing correction values were the same for all
samples. The 2D and 3D analysis was performed with CTAn software, provided by the device manufacturer. The same threshold levels were applied in binary image histograms, and regions of interest (ROIs) were determined manually. To measure and analyze various aspects of bone architecture, the following variables were studied: (1) BMD. (2) Bone volume/tissue volume (BV/TV) quantiﬁes bone quantity. It is measured by dividing the number of bone voxels of the volume of interest (VOI) by the total number of voxels in the VOI studied. This measurement is expressed as a percentage, and it is commonly used in pathological studies where bone remodeling is altered because it reﬂects bone gain or loss. (3) Trabecular thickness (Tb.Th) determines the average width of bone structure. This parameter is independent of the underlying model (rod or plate) and is measured by ﬁlling spheres into the analyzed structure. (4) Connectivity density (Conn.D) is the number of trabecular elements that can be removed without changing the bone network. It also estimates the number of trabecular connections per mm3. Bone trabeculae form closed loops, so connectivity density can also be considered as the quantiﬁcation of these closed loops in the trabecular bone net-
work. This topological parameter derives from the Euler number, and it is measured in (1-Euler number)/VOI and has been reported as the parameter most commonly affected by osteoporosis. Loss of bone trabeculae will generate a less connected network. (5) Trabecular number (Tb.N) is the number of traversals across bone trabeculae per unit of length, made on a random path through the VOI. This method consists on drawing a line through the ROI and counting the number of times it goes across the trabeculae. The mean value that establishes Tb.N is calculated by drawing an inﬁnite number of lines in all directions. (6) Trabecular separation (Tb.Sp) refers to the distance between trabeculae and corresponds to the bone marrow measurement. (7) Degree of anisotropy (DA) deﬁnes the asymmetry of the object analyzed and quantiﬁes the direction of its trabeculae. The bigger this number, the more aligned the trabecular bone will be. (8) Structure model index (SMI) indicates relative prevalence of rod-like or plate-like trabecular bone. It is deﬁned in an interval between 0 and 3, where 0 is an ideal plate-like structure and 3 is a cylinder.
RESULTS Table 2 shows the mean values for BMD expressed in grams of
Table 1. Patients’ Clinical Features Patient
Medical Condition Multiple myeloma Osteoporosis
Zoledronic acid 17 mo Alendronate 48 mo
Mandibular posterior Mandibular anterior
Implant loading Implant surgerytriggered
Management Sequestrectomy Sequestrectomy and PRGF
Complete resolution Complete resolution
PRGF indicates plasma rich in growth factors.
Table 2. Bone Mineral Density of the Samples Analyzed Measured in Grams per Cubic Centimeter BMD Control Case 1. Case 1. Case 2. Case 2.
Lower Upper Lower Upper
area area area area
Standard Error of the Mean
95% Conﬁdence Interval (Minimum)
95% Conﬁdence Interval (Maximum)
0.44355 0.32575 0.67357 0.86655 0.88905
0.43723 0.46786 0.55422 0.27804 0.29421
0.00089422 0.000088877 0.00021546 0.00012493 0.00017221
0.44177 0.32557 0.67314 0.8663 0.8887
0.44534 0.32593 0.674 0.8668 0.88939
IMPLANT DENTISTRY / VOLUME 23, NUMBER 3 2014
Table 3. Results of the BV/TV, Trabecular Thickness (Tb.Th), Trabecular Separation (Tb.Sp), and Trabecular Number (Tb.N) Control Case 1. Case 1. Case 2. Case 2.
Lower Upper Lower Upper
area area area area
32.865 29.024 53.057 66.844 66.402
0.12176 0.1677 0.19597 0.1575 0.16782
0.13838 0.38992 0.38384 0.12248 0.12848
2.699 1.7307 2.7074 4.244 3.9568
Fig. 7. Case 1. Three-dimensional reconstruction of the dental implant and the alveolar part of the analyzed bone.
Fig. 8. Case 2. Microarchitecture of necrotic bone in the basal region.
hydroxyapatite per cubic centimeter. The degree of mineralization observed in the osteonecrotic bone was higher than that observed in healthy bone. Furthermore, in bone sequestra, this value decreased while alveolar depth increased. Table 3 shows the data related to structural variables of our samples. Bone volume fraction BV/TV, measured as a percentage (%), represents the actual amount of bone in the studied area. This value is usually higher in osteonecrotic bone than in healthy bone. However, in long-term sequestrum associated with implant failure after 18 months of loading (case 1), BV/TV values obtained in the basal bone were signiﬁcantly lower than those in healthy bone or the rest of osteonecrotic areas of interest (Fig. 7). Trabecular thickness returned higher values in osteonecrotic bone than in healthy bone, with lower values in its lowest part. Trabecular separation was also slightly higher in osteonecrotic bone in the patient with longterm sequestrum but not in the patient with implant surgery-related osteonecrosis. Trabecular number in the patient with implant surgery-triggered osteonecrosis (case 2) was signiﬁcantly higher than that observed in the long-term sequestrum due to prosthetic loading and that of the control sample (Fig. 8). When analyzing DA (Table 4), we observed slightly low values in all sequestra samples as compared with
Table 4. Detailed Results for the SMI, DA, Euler Number (Eu.N), and Connective Density (ConnD) Control Case 1. Case 1. Case 2. Case 2.
Lower Upper Lower Upper
area area area area
3.774 2.0179 −2.6872 0.40661 1.4415
1.4564 2.935 1.3316 1.2955 1.3685
219 468 258 −882 −869
854.92 206.8 35.854 1258.9 1631.2
the control sample. However, the lower part in the alveolar process in long-term evolution osteonecrosis proved to be an exception. We also noticed that connective density values were signiﬁcantly high in implant surgery-triggered sequestrum but substantially low in sequestra caused by loading as compared with those of the control sample (Table 4). Euler number proved to be negatively correlated to this variable. Finally, Table 4 shows that SMI values observed in the control sample were closer to 3 (cylindrical rod structure) than those in osteonecrotic bone (plate-like structure).
DISCUSSION To our knowledge, this study is the ﬁrst analysis of bone microstructure in 2 patients with BRONJ associated with dental implants. Our aim was to clarify the characteristics of the bone around the implant in patients affected with BRONJ. Bone strength and resistance to fracture are determined by structural and material properties such as architecture, mineralization, and bone turnover.19,20 Changes in mineralization can alter bone tissue stiffness and strength, therefore inﬂuencing its mechanical properties. However, very high values of mineralization lead to higher bone fragility, resulting in very fragile bones prone to fracture.21 Thus, there is a trade-off between stiffness and toughness because an optimal balance in bone remodeling is needed to achieve a maximum resistance to fracture.19,22 In 1997, Müller and Rüegsegger23 investigated the 3D microstructure of trabecular bone using microCT, which led them to state that it is a combination of interconnecting thin rods and paralleloriented plates, as Singh24 had already described in 1978. In their study of 70 iliac crest bone biopsies, they observed a high correlation between histomorphometric values and BV/TV obtained using microCT.23 In 2004, Moon et al16 carried out a study on 3D jaw bone microstructure by analyzing bone fragments from 10 cadavers (7 male and 3 female bodies)
BRONJ, DENTAL IMPLANTS,
with ages ranging from 29 to 75 years old. Mean values for BV/TV were 43.74%, signiﬁcantly lower than those obtained in this study for bone sequestra in BRONJ patients. These authors divided the alveolar process in 4 different regions, 2 upper regions (alveolar bone) and 2 lower regions (basal bone), 1 above and 1 under the alveolodental ligament. Cadaveric jaw bone samples showed BV/TV and Tb.Th values that decreased whenever bone depth increased, as opposed to Tb.Sp and SMI, which increased as well. Several studies associated reduced trabecular thickness and rod-like geometry with severe damage in bone formation.22 In 2012, González-García and Monje25 used microCT to study 39 human alveolar bone biopsies with dental implants, stating a correlation between microCT results and histomorphometry. These authors did not observe any signiﬁcant differences for the following variables: age (ranging from 20 to 79 years old), gender, and bone location. Even though 66% of the bone biopsies were from the upper jaw, the mean value for BV/TV was of 48.7%, also lower than that observed for the sequestra in our study. They observed a signiﬁcant correlation between BV/ TV, thickness, and trabecular number (the higher the BV/TV, the higher the Tb.Th and Tb.N) and between BV/TV, trabecular separation, and SMI (the higher the BV/TV, the lower the Tb.Sp and SMI). In our study, trabecular separation ranged from 0.12 to 0.38 mm, similar to the values deﬁned as normal in other studies (0.31 and 0.51 mm).16,25 Regarding SMI, values observed for the control sample were closer to 3 (cylindrical rod structure) than those in bone sequestra (plate-like structure). The SMI parameter was introduced to quantify the 3D morphology of trabecular bone in terms of plates and rods in the structure. SMI measures the convexity of the trabecular surface through dilation of the 3D model. Concave surfaces in closed bone cavities represent a negative SMI value, so bone regions with many closed cavities (usually with a high BV/TV) may have
negative SMI values. As a consequence, several studies have determined a relationship between SMI and BV/TV.25 Although recent studies do not include Conn.D and Euler number as variables, Feldkamp et al’s26 classic study refers to their signiﬁcant correlation with BV/TV and trabecular number. After analyzing vertebral specimens of male and female minipigs of different age, other authors27 determine that Conn.D can distinguish trabecular structure in young minipigs (higher value) from mature minipigs (lower value). According to them, even though 74% to 85% of bone resistance and elasticity depend on mineral density and BV/TV, other variables such as Conn.D may entrain bone resistance capacity. In our study, Conn.D had extremely high values in implant surgery-triggered sequestrum (case 2) as opposed to the values in sequestrum due to loading before BP administration (case 1), which were signiﬁcantly lower than those in the control sample. These observations may explain the model of BRONJ associated with dental implants proposed by Kwon et al.12 In that study, it is described the 3 types of bone destruction pattern in patients suffering from osteonecrosis associated with dental implants: frozen type (complete necrosis of the bone around the implant), osteolytic type (extensive osteolysis around the implant with or without sequestra), and en block sequestration type (sequestration of bone with an implant maintaining direct implant-bone contact). These patterns could exist simultaneously depending on the degree of local bone destruction and the severity of infection. Figures 4 and 8 revealed alveolar bone around the implant is completely necrotic, which leads to a loss of stability. Progressive osteolysis results in sequestration. In Figures 3 and 7, en block sequestration includes the implant, with osteolysis and infection also affecting nonintegrated bone. Given the limitation inherent in our study, which is the reduced number of cases, this observation could also explain the presence of spontaneous BRONJ (without previous exposure to
dental surgery). Osteonecrosis related to dental extraction would be present in patients with high Conn.D values. However, 20% to 30% of osteonecrosis deﬁned as spontaneous would appear in bone with very low Conn.D values. At present, information on microCT analysis of bone sequestrum in patients with BRONJ is limited. Allen and Ruggiero28 analyzed bone sequestra in 13 patients with BRONJ using microCT and comparing the results to 8 cadaveric jaw bone samples. They studied 2 variables: the mean value of the grayscale and the percentage of mineralization per region. They observed that sequestra in BRONJ patients had 9% more mineralized tissue. In 2010, the same authors29 investigated BV/TV values in sequestra from 31 BRONJ patients using microCT. The aim of this study was to evaluate if there were differences between sequestra regarding the time of evolution of the osteonecrosis. Although BV/TV values were high, no signiﬁcant differences were observed between the sequestra in terms of time of evolution. In our study, BV/TV values did not show relevant differences either.
CONCLUSIONS Despite the signiﬁcant limitation related to the reduced number of sequestra analyzed in this study and also the control sample being from a healthy patient instead of a patient in BP treatment with no osteonecrosis, we can conclude that direct and indirect characteristics of BRONJ bone sequestra associated with dental implants show an increase in BV/TV, trabecular number, and trabecular thickness with a decrease in trabecular separation. Conn.D also increased and SMI tends to the ideal plate-like model. However, more studies will be needed to determine if these values depend on the time of evolution or on the mechanism of the osteonecrosis.
DISCLOSURE The authors claim to have no ﬁnancial interest, either directly or indirectly, in the products or information listed in the article.
IMPLANT DENTISTRY / VOLUME 23, NUMBER 3 2014
REFERENCES 1. Merigo E, Manfredi M, Meleti M, et al. Bone necrosis of the jaws associated with bisphosphonate treatment: A report of twenty-nine cases. Acta Biomed. 2006;77:109–117. 2. Gutta R, Louis PJ. Bisphosphonates and osteonecrosis of the jaws: Science and rationale. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;104: 186–193. 3. Donnelly E, Chen DX, Boskey AL, et al. Contribution of mineral to bone structural behavior and tissue mechanical properties. Calcif Tissue Int. 2010;87:450–460. 4. Haworth AE, Webb J. Skeletal complications of bisphosphonate use: What the radiologist should know. Br J Radiol. 2012;85:1333–1342. 5. Masoodi NA. Oral bisphosphonates and the risk for osteonecrosis of the jaw. Br J Med Pract. 2009;2:11–15. 6. Marx RE. Pamidronate (Aredia) and zoledronate (Zometa) induced avascular necrosis of the jaws: A growing epidemic. J Oral Maxillofac Surg. 2003;61:1115– 1117. 7. Junquera L, Gallego L, Cuesta P, et al. Clinical experiences with bisphosphonateassociated osteonecrosis of the jaws: Analysis of 21 cases. Am J Otolaryngol. 2009;30:390–395. 8. Sivolella S, Lumachi F, Stellini E, et al. Denosumab and anti-angiogenetic drug-related osteonecrosis of the jaw: An uncommon but potentially severe disease. Anticancer Res. 2013;33:1793–1797. 9. Bedogni A, Bettini G, Totola A, et al. Oral bisphosphonate-associated osteonecrosis of the jaw after implant surgery: A case report and literature review. J Oral Maxillofac Surg. 2010;68:1662–1666. 10. Lazarovici TS, Yahalom R, Taicher S, et al. Bisphosphonate-related osteonecrosis of the jaws: A single-center study of 101 patients. J Oral Maxillofac Surg. 2009;67:850–855. 11. Favia G, Piattelli A, Sportelli P, et al. Osteonecrosis of the posterior mandible after implant insertion: A clinical and histo-
logical case report. Clin Implant Dent Relat Res. 2011;13:58–63. 12. Kwon TG, Lee CO, Park JW, et al. Osteonecrosis associated with dental implants in patients undergoing bisphosphonate treatment. Clin Oral Impl Res. (In Press). 13. Tam Y, Kar K, Nowzari H, et al. Osteonecrosis of the jaw after implant surgery in patients treated with bisphosphonates: a presentation of six consecutive cases. Clin Implant Dent Relat Res. (In Press). 14. Jacobsen C, Metzler P, Rössle M, et al. Osteopathology induced by bisphosphonates and dental implants: Clinical observations. Clin Oral Investig. 2013; 17:167–175. 15. Müller R, Hahn M, Vogel M, et al. Morphometric analysis of noninvasively assessed bone biopsies: Comparison of high-resolution computed tomography and histologic sections. Bone. 1996;18: 215–220. 16. Moon HS, Won YY, Kim KD, et al. The three-dimensional microstructure of the trabecular bone in the mandible. Surg Radiol Anat. 2004;26:466–473. 17. Tamminen IS, Isaksson H, Aula AS, et al. Reproducibility and agreement of micro-CT and histomorphometry in human trabecular bone with different metabolic status. J Bone Miner Metab. 2011; 29:442–448. 18. Fanuscu MI, Chang TL. Threedimensional morphometric analysis of human cadaver bone: Microstructural data from maxilla and mandible. Clin Oral Implants Res. 2004;15:213–228. 19. Borah B, Dufresne TE, Ritman EL, et al. Long-term risedronate treatment normalizes mineralization and continues to preserve trabecular architecture: Sequential triple biopsy studies with micro-computed tomography. Bone. 2006;39:345–352. 20. Felsenberg D, Boonen S. The bone quality framework: Determinants of bone strength and their interrelationships, and implications for osteoporosis management. Clin Ther. 2005;27:1–11.
21. Ciarelli TE, Fyhrie DP, Parﬁtt AM. Effects of vertebral bone fragility and bone formation rate on the mineralization levels of cancellous bone from white females. Bone. 2003;32:311–315. 22. Green JO, Diab T, Allen MR, et al. Three years of alendronate treatment does not continue to decrease microstructural stresses and strains associated with trabecular microdamage initiation beyond those at 1 year. Osteoporos Int. 2012;23: 2313–2320. 23. Müller R, Rüegsegger P. Microtomographic imaging for the nondestructive evaluation of trabecular bone architecture. Stud Health Technol Inform. 1997;40:61–79. 24. Singh I. The architecture of cancellous bone. J Anat. 1978;127:305–310. 25. González-García R, Monje F. Is micro-computed tomography reliable to determine the microstructure of the maxillary alveolar bone? Clin Oral Implants Res. 2013;24:730–737. 26. Feldkamp LA, Goldstein SA, Parﬁtt AM, et al. The direct examination of threedimensional bone architecture in vitro by computed tomography. J Bone Miner Res. 1989;4:3–11. 27. Borah B, Dufresne TE, Cockman MD, et al. Evaluation of changes in trabecular bone architecture and mechanical properties of minipig vertebrae by three-dimensional magnetic resonance microimaging and ﬁnite element modeling. J Bone Miner Res. 2000;15: 1786–1797. 28. Allen MR, Ruggiero SL. Higher bone matrix density exists in only a subset of patients with bisphosphonate-related osteonecrosis of the jaw. J Oral Maxillofac Surg. 2009;67:1373–1377. 29. Allen MR, Pandya B, Ruggiero SL. Lack of correlation between duration of osteonecrosis of the jaw and sequestra tissue morphology: What it tells us about the condition and what it means for future studies. J Oral Maxillofac Surg. 2010;68: 2730–2734.