- Email: [email protected]
Evaluation of the mechanical and architectural properties of glenoid bone
Evaluation of the mechanical and architectural properties of glenoid bone Raghad Mimar, PhD,a David Limb, BSc FRCS Ed(Orth),b and Richard M. Hall, PhD,a,c Leeds, United Kingdom
Successful glenoid fixation in shoulder arthroplasty is partly dependent on the properties of the underlying bone. Therefore, mapping of the glenoid surface and locating the bone with the highest quality, in terms of mechanical properties and morphology, is a key requirement in ensuring effective fixation. To this end, an investigation was undertaken to study the relationship between indentation behavior and the quality of the glenoid bone. Nineteen embalmed glenoids were obtained from human cadavers (mean age at death, 82 years). Each specimen was tested using a cylindrical indentor at 11 predetermined points to investigate loaddisplacement behavior. Microcomputed tomography analysis was performed to ascertain the bone volume (BV)/total volume (TV) fraction of the trabecular bone and the subchondral thickness. Statistical analysis showed that both strength and modulus varied with indentation position. Significant relationships were found between either strength or modulus and BV/TV or subchondral thickness, although the explained variance was relatively low. (J Shoulder Elbow Surg 2008;17:336-341.)
Shoulder arthroplasty has been used successfully in
patients with shoulder arthritis since the 1960s. It is designed to replace the diseased joint, provide enhanced function, and improve the quality of life; however, loosening of the glenoid component is a significant long-term complication in total shoulder arthroplasty.9,11,13 This loosening process is multifactorial, with a number of contributory factors that encompass both mechanical and biological attributes. These factors include the amount of cement used in From the aSchool of Mechanical Engineering, University of Leeds; b Department of Orthopaedic Surgery, Leeds Teaching Hospitals Trust; and cAcademic Unit of Orthopaedic Surgery, Leeds General Infirmary. This work was supported by the Iranian Ministry of Science Studentship. Reprint requests: Professor Richard M. Hall, School of Mechanical Engineering, University of Leeds, Leeds LS2 9JT, UK (E-mail: R. [email protected]). Copyright ª 2008 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2008/$34.00 doi:10.1016/j.jse.2007.07.024
336
the fixation and the risk of thermal injury to the bone,4 eccentric loading of the glenoid component due to translation of the humeral head,3 and poor cementing technique.18 Another critical variable in successful glenoid component fixation is the quality of the underlying bone together with the level of interdigitation of cement with the trabecular structure.1 When the fixation deteriorates, this supporting bone maybe eroded further, resulting in cavitation, more difficult revision procedures, and poorer outcomes. Thus, knowledge of the bone’s architecture and structural properties, including spatial variations, are of great importance in deducing the most effective fixation for glenoid components in total shoulder arthroplasty. A number of methods are available for investigating the mechanical properties and bony architecture in the glenoid.1,6,8,10,15 For the former, indentation testing has been used previously to determine regional changes in the mechanical properties of the glenoid surface,1 and histologic or microcomputed tomography (micro-CT) techniques have been used to deduce bone morphology.2,17 However, little has been published that relates mechanical properties to the quality or structure of the surrounding bone. The aim of this study was to investigate the relationship between the mechanical properties of the glenoid surface and both the trabecular bone volume (BV)/total volume (TV) fraction and the subchondral thickness of the glenoid surface. This study defines subchondral thickness as the thickness of the subchondral plate of bone from the surface to the first manifestation of trabecular spaces. Both BV/TV and the subchondral thickness are predicted to affect the local strength and stiffness of bone in the glenoid fossa. MATERIALS AND METHODS Mechanical properties Nineteen glenoids were obtained from embalmed cadavers of 16 women and 3 men (mean age, 82 years) who did not have macroscopic evidence of structural or degenerative changes in the rotator cuff or articular surfaces of the shoulder as determined by direct visual inspection. Anglin et al1 have previously indicated that embalming has minimal impact on the mechanical properties of bone
Mimar, Limb, and Hall
J Shoulder Elbow Surg Volume 17, Number 2
337
Figure 1 Preselected indentation points on the glenoid surface are shown. Position of the points was selected relative to the maximum width and length of the glenoid.
derived using indentation testing.16 Soft tissues and muscle attachments were removed from the acromion and coracoid by sharp dissection, as was the glenoid labrum, leaving the bony glenoid fossa with its articular cartilage intact. The articular cartilage was carefully removed by immersing the glenoid fossa in a solution of papain (Sigma-Aldrich, Stenheim, Germany) at 40 C for 6 to 8 weeks. The cartilage surface was scored to facilitate penetration of the papain and aid disintegration and removal. The structural properties of the glenoid surface were tested using an indentor technique based on that previously utilized by Anglin et al.1 A 2.5-mm flat cylindrical indentor was attached to a tensile testing machine (Shimadzu Autograph, Tokyo, Japan) that penetrated the glenoid bone at predetermined positions on a grid of 11 points. The location of the test sites was mapped in relation to proportional positions along the glenoid width and length to standardize the results (Figure 1). The specimens were mounted in a polymethylmethacrylate mold that itself was located on an angle jig that could be adjusted to ensure that the indentor was positioned normally to the surface for all test measurements, regardless of surface curvature (Figure 2). Indentation was performed at rate of 2 mm/min to a depth of 3 mm from the initial point of contact with the glenoid surface. The indentation process was continually monitored, allowing the derivation of a load-deformation curve. The strength of the bone for each indentation site
Figure 2 An angle jig was used to set the alignment to obtain a normal force to each indentation site.
was calculated as the failure load divided by the area of the indentor.1 The modulus was calculated using the equation: E ¼ ½1 v2 =2r ,½P=d,19 which has been used previously in these types of experiments,1 where d is the Poisson ratio, 2r is the diameter of the indentor, and P/d is the maximum slope determined from the load-deformation curve. A value of 0.25 was allocated to the Poisson ratio, as this has previously been used in this type of study.12,16
Bone architecture The architecture of the trabecular bone and the subchondral thickness at the glenoid surface was investigated using the micro-CT technique. The micro-CT (mCT 80 SCANCO Medical AG, Switzerland) scanner allowed noninvasive scanning without damage to the glenoid specimen. The resolution used was 78 mm, allowing the observation of all but the most slender trabeculae. Figure 3 displays a section through a 3-dimensional (3D) representation of the glenoid bone architecture. The trabecular bone of the glenoid was digitally separated from its cortical shell by manually outlining the contour of the cortical-cancellous boundary in each
338
Mimar, Limb, and Hall
J Shoulder Elbow Surg March/April 2008
Figure 3 This section of the three-dimensional representation of glenoid bone architecture was created by using microcomputed tomography.
Figure 4 Graph shows the load-deformation behavior for a typical indentation test. The elastic modulus was determined from the steepest portion of the rising curve, whereas the strength was determined from the maximum load sustained by the bone.
Table I Mean values for the strength, elastic modulus, subchondral thickness, and bone volume fraction of the cancellous bone for each of the 11 points identified in Figure 1
Zone 1 2 3 4 5 6 7 8 9 10 11
Strength, mean ± SD MPa
Elastic modulus, mean ± SD MPa
BV/TV, mean ± SD %
Cortical thickness, mean ± SD mm
39 6 12 57 6 16 56 6 17 34 6 13 26 6 10 67 6 30 44 6 15 28 6 08 48 6 15 58 6 16 33 6 15
154 6 49 179 6 42 234 6 66 148 6 61 119 6 38 182 6 73 189 6 81 123 6 41 144 6 44 210 6 52 146 6 59
0.15 6 0.08 0.24 6 0.09 0.22 6 0.11 0.14 6 0.05 0.16 6 0.07 0.24 6 0.11 0.28 6 0.07 0.19 6 0.07 0.24 6 0.10 0.41 6 0.08 0.27 6 0.08
0.95 6 0.19 1.04 6 0.18 1.12 6 0.15 0.95 6 0.19 0.88 6 0.21 1.11 6 0.23 1.15 6 0.19 0.93 6 0.18 1.11 6 0.18 1.09 6 0.19 1.01 6 0.26
BV, Bone volume; TV, total volume.
specimen. Morphologic parameters of the cancellous bone could then be calculated using a cylindrical subvolume of the original image, referred to as the volume of interest (VOI). For each deformation zone generated during indentation, a VOI was chosen immediately below the cancellous bone compressed by the indentor, the boundaries of which could be visually identified on 2D micro-CT images. The VOI in each case measured 63 mm3. The bone volume fraction (BV/TV) was determined for each VOI. The thickness of the subchondral bone was assessed separately by adjusting the plane of the images of the indented glenoid to exactly 90 to the surface at the indentation point (Figure 3). Two-dimensional images in this plane allowed measurement of subchondral thickness at either side of the indentation site. By spinning this plane through 90 around the axis of the indentation channel, a second set of measurements could be obtained so that the subchondral thickness adjacent to the hole at four equally separated points could be measured manually. The average of the 4 points was determined to give a mean thickness value. Although the
calcified chondral layer makes determination of subchondral thickness difficult in fresh bone, this is not applicable to cadaveric bones prepared in papain, which have an easily defined surface plane. Summary statistics were determined for each of the parameters for each glenoid and indentation sites. Linear regression analysis was performed to assess the relationship between the elastic modulus and strength with the subchondral thickness and bone volume fraction. The calculations were done using the STATA 9 computer program (StataCorp, College Station, TX). Significance was set at P ¼ .05.
RESULTS Presented in Figure 4 is a typical load-displacement curve for an indentation test on the glenoid surface. All the plots were characterized by a small toe region, followed by a relatively linear portion of the curve. The elastic modulus was determined from the
J Shoulder Elbow Surg Volume 17, Number 2
Mimar, Limb, and Hall
339
Figure 5 Photographs show the variation in the strength, elastic modulus, bone volume to total volume ratio, and subchondral thickness across the glenoid surface. The red dots indicate indentation sites in which the parameters were at the median value or higher. It can be clearly observed that that the lowest values and, hence, the poorest quality bone were found in the inferior portion of the glenoid.
maximum tangent gradient on this segment of the curve. The strength of the bone was taken to be the maximum load observed in the load-deformation curve divided by the cross-sectional area of the indentor. In those instances in which more than 1 peak was observed, this maximum was defined to be the first maximum value, followed by a drop in load of 10 % or more. The mean strength for each of the indentation sites was 26 to 67 MPa (Table I). One-way analysis of
variance indicated that there were significant differences in the strength of the bone among positions on the glenoid (F ¼ 14.0, P < .001). Similar results were observed for the elastic modulus, in which the mean value was 119 to 234 MPa, again, with a significant variation with position across the glenoid (F ¼ 7.6, P < .001). In general, there was a trend for the larger values of BV/TV to be found at the posterior margins of the glenoid, whereas the more porous
340
Mimar, Limb, and Hall
regions were observed near the anteroinferior edge. The differences in bone volume fraction across the glenoid were found to be significant (F ¼ 14.5, P <.001). The measured subchondral thickness was 0.88 mm to 1.15 mm, with thinner bone located inferiorly; again, analysis of variance demonstrated that the differences with position were significant. Multiple regression analysis, in which account was taken of the lack of independence of results taken from the same glenoid, indicated that both the bone volume fraction and the subchondral thickness have a significant effect on the elastic modulus. Similar results were observed for the relationship between strength and bone volume fraction and subchondral thickness. However, the R2 values of 0.21 for the elastic modulus and 0.31 for the strength analyses indicate that these linear models are poor and that other factors may play a significant role in determining the mechanical properties of the glenoid surface. DISCUSSION Regional differences were found in both the mechanical parameters and the morphologic measures, in keeping with previous studies.1,7 However, there were disparities in the absolute values of the parameters measured compared with those observed elsewhere. Frich et al7 reported mean ultimate strengths of approximately 35 MPa to more than 110 MPa, whereas the results reported here are roughly half these values. One possible reason for this discrepancy is the use of much younger specimens in the Frich et al11 cohort, which had a mean age of younger than 60 years. The results reported by Anglin et al1 were somewhat lower than those cited here or elsewhere and arise principally from the different methodology used in that multiple readings at different depths were taken, which included results from the weaker trabecular bone. Further disparities in both the elastic modulus and strength reported here and those accrued elsewhere can result from a number of other factors, including dissimilar specimens preparations, the removal of the subchondral bone,1 and significant variation in the experimental protocols and analyses.6 From this study, we believe that age and the use of female specimens are the most likely explanation for the thinner subchondral thickness we observed compared with past studies.5,7 To our knowledge, this is the first study to combine indentation and micro-CT analyses to investigate the relationship between the glenoid’s mechanical properties and bone morphology. The distribution of BV/ TV data is similar to that found for bone mineral density measurements by Lehtinen et al.14 However, the subchondral thickness measurements differ substantially from those previously reported7 and may result from differences in specimen preparation.
J Shoulder Elbow Surg March/April 2008
Both the local bone volume fraction and the subchondral thickness influenced the strength at the glenoid surface as well as the elastic modulus of the bone. The explained variance was small, however, and other factors not incorporated into this model may also have a significant effect on the strength and elastic modulus. These factors could include the degree of mineralization and architectural indices such as the anisotropy. In this particular set of experiments, embalmed cadaveric tissue was used. Anglin et al1 had previously suggested that this preservation had minimal impact on the quality of the results for this type of experiment, although it is difficult to compare the data reported here and those cited elsewhere to verify this assertion. If the embalming process were found to modify the mechanical properties of the tissue, then, provided this occurred in a relatively straightforward fashion, the relationships between strength or stiffness and the morphologic parameters would still likely hold true. Current work with fresh cadaveric tissue is ongoing to explore these issues further. Currently, the study has possibly identified that there is a band of bone across the glenoid that appears to provide superior mechanical properties for fixation according to the parameters we have identified. This runs from the base of the coracoid backward through the central glenoid to the posterior margin. This region should provide the best support for a glenoid component, although results will have to be verified on fresh bone. Previous studies by Frich6 have identified similar regional variation in bone properties, in which it was reported that the strongest bone lay posterior and superior, as well as anterior, to the bare area. Our study agrees with that of Frich in recognizing the poorer mechanical properties of bone at the inferior parts of the glenoid (Figure 5). It is notable that the anteroinferior region has the poorest structural and biomechanical properties. Interestingly, this lies within the region observed clinically that has a propensity for bone loss due to instability. This relationship merits further study. In summary, this study has identified a very useful combination of tools for studying this anatomic region that will facilitate development of more reliable glenoid fixation. We thank Dr Roger Soames for all of his contributions to this project.
REFERENCES
1. Anglin C, Tolhurst P, Wyss UP, Pichora DR. Glenoid cancellous bone strength and modulus. J Biomech 1999;32:1091-7. 2. Boyd SK, Muller R, Matyas JR, Wohl GR, Zernicke RF. Early morphometric and anisotropic change in periarticular cancellous bone in a model of experimental knee osteoarthritis quantified using microcomputed tomography. Clin Biomech 2000;15: 624-31.
J Shoulder Elbow Surg Volume 17, Number 2
3. Buchler P, Ramaniraka NA, Rakotomanana LR, Iannotti JP, Farron A. A finite element model of the shoulder: application to the comparison of normal and osteoarthritic joints. Clin Biomech 2002;17:630-9. 4. Churchill RS, Boorman RS, Fehringer EV, Matsen FA 3rd. Glenoid cementing may generate sufficient heat to endanger the surrounding bone. Clin Orthop Relat Res 2004:76-9. 5. Eckstein F, Milz S, Anetzberger H, Putz R. Thickness of the subchondral mineralised tissue zone (SMZ) in normal male and female and pathological human patellae. J Anat 1998;192:81-90. 6. Frich LH, Jensen NC, Odgaard A, Pedersen CM, Sojbjerg JO, Dalstra M. Bone strength and material properties of the glenoid. J Shoulder Elbow Surg 1997;6:97-104. 7. Frich LH, Odgaard A, Dalstra M. Glenoid bone architecture. J Shoulder Elbow Surg 1998;7:356-61. 8. Goldstein SA. The mechanical properties of trabecular bone: dependence on anatomic location and function. J Biomech 1987;20:1055-61. 9. Hasan SS, Leith JM, Campbell B, Kapil R, Smith KL, Matsen FA 3rd. Characteristics of unsatisfactory shoulder arthroplasties. J Shoulder Elbow Surg 2002;11:431-41. 10. Hvid I, Rasmussen O, Jensen NC, Nielsen S. Trabecular bone strength profiles at the ankle joint. Clin Orthop Relat Res 1985:306-12. 11. Ibarra C, Dines DM, McLaughlin JA. Glenoid replacement in total shoulder arthroplasty. Orthop Clin North Am 1998;29: 403-13.
Mimar, Limb, and Hall
341
12. Keaveny TM, Hayes WC. A 20-year perspective on the mechanical properties of trabecular bone. J Biomech Eng 1993; 115:534-42. 13. Lacroix D, Murphy LA, Prendergast PJ. Three-dimensional finite element analysis of glenoid replacement prostheses: a comparison of keeled and pegged anchorage systems. J Biomech Eng 2000; 122:430-6. 14. Lehtinen JT, Tingart MJ, Apreleva M, Warner JJ. Total, trabecular, and cortical bone mineral density in different regions of the glenoid. J Shoulder Elbow Surg 2004;13:344-8. 15. Linde F, Pongsoipetch B, Frich LH, Hvid I. Three-axial strain controlled testing applied to bone specimens from the proximal tibial epiphysis. J Biomech 1990;23:1167-72. 16. Linde F. Elastic and viscoelastic properties of trabecular bone by a compression testing approach. Dan Med Bull 1994;41: 119-38. 17. Mu¨ller R, Hahn M, Vogel M, Delling G, Ru¨egsegger P. Morphometric analysis of noninvasively assessed bone biopsies: comparison of high-resolution CT and histologic sections. Bone 1996;18: 215-20. 18. Terrier A, Buchler P, Farron A. Bone-cement interface of the glenoid component: stress analysis for varying cement thickness. Clin Biomech 2005;20:710-7. 19. Timoshenko SP, Goodier JN. Theory of elasticity. 3rd ed. New York, NY: McGraw-Hill; 1970. p. 408.
Successful glenoid fixation in shoulder arthroplasty is partly dependent on the properties of the underlying bone. Therefore, mapping of the glenoid surface and locating the bone with the highest quality, in terms of mechanical properties and morphology, is a key requirement in ensuring effective fixation. To this end, an investigation was undertaken to study the relationship between indentation behavior and the quality of the glenoid bone. Nineteen embalmed glenoids were obtained from human cadavers (mean age at death, 82 years). Each specimen was tested using a cylindrical indentor at 11 predetermined points to investigate loaddisplacement behavior. Microcomputed tomography analysis was performed to ascertain the bone volume (BV)/total volume (TV) fraction of the trabecular bone and the subchondral thickness. Statistical analysis showed that both strength and modulus varied with indentation position. Significant relationships were found between either strength or modulus and BV/TV or subchondral thickness, although the explained variance was relatively low. (J Shoulder Elbow Surg 2008;17:336-341.)
Shoulder arthroplasty has been used successfully in
patients with shoulder arthritis since the 1960s. It is designed to replace the diseased joint, provide enhanced function, and improve the quality of life; however, loosening of the glenoid component is a significant long-term complication in total shoulder arthroplasty.9,11,13 This loosening process is multifactorial, with a number of contributory factors that encompass both mechanical and biological attributes. These factors include the amount of cement used in From the aSchool of Mechanical Engineering, University of Leeds; b Department of Orthopaedic Surgery, Leeds Teaching Hospitals Trust; and cAcademic Unit of Orthopaedic Surgery, Leeds General Infirmary. This work was supported by the Iranian Ministry of Science Studentship. Reprint requests: Professor Richard M. Hall, School of Mechanical Engineering, University of Leeds, Leeds LS2 9JT, UK (E-mail: R. [email protected]). Copyright ª 2008 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2008/$34.00 doi:10.1016/j.jse.2007.07.024
336
the fixation and the risk of thermal injury to the bone,4 eccentric loading of the glenoid component due to translation of the humeral head,3 and poor cementing technique.18 Another critical variable in successful glenoid component fixation is the quality of the underlying bone together with the level of interdigitation of cement with the trabecular structure.1 When the fixation deteriorates, this supporting bone maybe eroded further, resulting in cavitation, more difficult revision procedures, and poorer outcomes. Thus, knowledge of the bone’s architecture and structural properties, including spatial variations, are of great importance in deducing the most effective fixation for glenoid components in total shoulder arthroplasty. A number of methods are available for investigating the mechanical properties and bony architecture in the glenoid.1,6,8,10,15 For the former, indentation testing has been used previously to determine regional changes in the mechanical properties of the glenoid surface,1 and histologic or microcomputed tomography (micro-CT) techniques have been used to deduce bone morphology.2,17 However, little has been published that relates mechanical properties to the quality or structure of the surrounding bone. The aim of this study was to investigate the relationship between the mechanical properties of the glenoid surface and both the trabecular bone volume (BV)/total volume (TV) fraction and the subchondral thickness of the glenoid surface. This study defines subchondral thickness as the thickness of the subchondral plate of bone from the surface to the first manifestation of trabecular spaces. Both BV/TV and the subchondral thickness are predicted to affect the local strength and stiffness of bone in the glenoid fossa. MATERIALS AND METHODS Mechanical properties Nineteen glenoids were obtained from embalmed cadavers of 16 women and 3 men (mean age, 82 years) who did not have macroscopic evidence of structural or degenerative changes in the rotator cuff or articular surfaces of the shoulder as determined by direct visual inspection. Anglin et al1 have previously indicated that embalming has minimal impact on the mechanical properties of bone
Mimar, Limb, and Hall
J Shoulder Elbow Surg Volume 17, Number 2
337
Figure 1 Preselected indentation points on the glenoid surface are shown. Position of the points was selected relative to the maximum width and length of the glenoid.
derived using indentation testing.16 Soft tissues and muscle attachments were removed from the acromion and coracoid by sharp dissection, as was the glenoid labrum, leaving the bony glenoid fossa with its articular cartilage intact. The articular cartilage was carefully removed by immersing the glenoid fossa in a solution of papain (Sigma-Aldrich, Stenheim, Germany) at 40 C for 6 to 8 weeks. The cartilage surface was scored to facilitate penetration of the papain and aid disintegration and removal. The structural properties of the glenoid surface were tested using an indentor technique based on that previously utilized by Anglin et al.1 A 2.5-mm flat cylindrical indentor was attached to a tensile testing machine (Shimadzu Autograph, Tokyo, Japan) that penetrated the glenoid bone at predetermined positions on a grid of 11 points. The location of the test sites was mapped in relation to proportional positions along the glenoid width and length to standardize the results (Figure 1). The specimens were mounted in a polymethylmethacrylate mold that itself was located on an angle jig that could be adjusted to ensure that the indentor was positioned normally to the surface for all test measurements, regardless of surface curvature (Figure 2). Indentation was performed at rate of 2 mm/min to a depth of 3 mm from the initial point of contact with the glenoid surface. The indentation process was continually monitored, allowing the derivation of a load-deformation curve. The strength of the bone for each indentation site
Figure 2 An angle jig was used to set the alignment to obtain a normal force to each indentation site.
was calculated as the failure load divided by the area of the indentor.1 The modulus was calculated using the equation: E ¼ ½1 v2 =2r ,½P=d,19 which has been used previously in these types of experiments,1 where d is the Poisson ratio, 2r is the diameter of the indentor, and P/d is the maximum slope determined from the load-deformation curve. A value of 0.25 was allocated to the Poisson ratio, as this has previously been used in this type of study.12,16
Bone architecture The architecture of the trabecular bone and the subchondral thickness at the glenoid surface was investigated using the micro-CT technique. The micro-CT (mCT 80 SCANCO Medical AG, Switzerland) scanner allowed noninvasive scanning without damage to the glenoid specimen. The resolution used was 78 mm, allowing the observation of all but the most slender trabeculae. Figure 3 displays a section through a 3-dimensional (3D) representation of the glenoid bone architecture. The trabecular bone of the glenoid was digitally separated from its cortical shell by manually outlining the contour of the cortical-cancellous boundary in each
338
Mimar, Limb, and Hall
J Shoulder Elbow Surg March/April 2008
Figure 3 This section of the three-dimensional representation of glenoid bone architecture was created by using microcomputed tomography.
Figure 4 Graph shows the load-deformation behavior for a typical indentation test. The elastic modulus was determined from the steepest portion of the rising curve, whereas the strength was determined from the maximum load sustained by the bone.
Table I Mean values for the strength, elastic modulus, subchondral thickness, and bone volume fraction of the cancellous bone for each of the 11 points identified in Figure 1
Zone 1 2 3 4 5 6 7 8 9 10 11
Strength, mean ± SD MPa
Elastic modulus, mean ± SD MPa
BV/TV, mean ± SD %
Cortical thickness, mean ± SD mm
39 6 12 57 6 16 56 6 17 34 6 13 26 6 10 67 6 30 44 6 15 28 6 08 48 6 15 58 6 16 33 6 15
154 6 49 179 6 42 234 6 66 148 6 61 119 6 38 182 6 73 189 6 81 123 6 41 144 6 44 210 6 52 146 6 59
0.15 6 0.08 0.24 6 0.09 0.22 6 0.11 0.14 6 0.05 0.16 6 0.07 0.24 6 0.11 0.28 6 0.07 0.19 6 0.07 0.24 6 0.10 0.41 6 0.08 0.27 6 0.08
0.95 6 0.19 1.04 6 0.18 1.12 6 0.15 0.95 6 0.19 0.88 6 0.21 1.11 6 0.23 1.15 6 0.19 0.93 6 0.18 1.11 6 0.18 1.09 6 0.19 1.01 6 0.26
BV, Bone volume; TV, total volume.
specimen. Morphologic parameters of the cancellous bone could then be calculated using a cylindrical subvolume of the original image, referred to as the volume of interest (VOI). For each deformation zone generated during indentation, a VOI was chosen immediately below the cancellous bone compressed by the indentor, the boundaries of which could be visually identified on 2D micro-CT images. The VOI in each case measured 63 mm3. The bone volume fraction (BV/TV) was determined for each VOI. The thickness of the subchondral bone was assessed separately by adjusting the plane of the images of the indented glenoid to exactly 90 to the surface at the indentation point (Figure 3). Two-dimensional images in this plane allowed measurement of subchondral thickness at either side of the indentation site. By spinning this plane through 90 around the axis of the indentation channel, a second set of measurements could be obtained so that the subchondral thickness adjacent to the hole at four equally separated points could be measured manually. The average of the 4 points was determined to give a mean thickness value. Although the
calcified chondral layer makes determination of subchondral thickness difficult in fresh bone, this is not applicable to cadaveric bones prepared in papain, which have an easily defined surface plane. Summary statistics were determined for each of the parameters for each glenoid and indentation sites. Linear regression analysis was performed to assess the relationship between the elastic modulus and strength with the subchondral thickness and bone volume fraction. The calculations were done using the STATA 9 computer program (StataCorp, College Station, TX). Significance was set at P ¼ .05.
RESULTS Presented in Figure 4 is a typical load-displacement curve for an indentation test on the glenoid surface. All the plots were characterized by a small toe region, followed by a relatively linear portion of the curve. The elastic modulus was determined from the
J Shoulder Elbow Surg Volume 17, Number 2
Mimar, Limb, and Hall
339
Figure 5 Photographs show the variation in the strength, elastic modulus, bone volume to total volume ratio, and subchondral thickness across the glenoid surface. The red dots indicate indentation sites in which the parameters were at the median value or higher. It can be clearly observed that that the lowest values and, hence, the poorest quality bone were found in the inferior portion of the glenoid.
maximum tangent gradient on this segment of the curve. The strength of the bone was taken to be the maximum load observed in the load-deformation curve divided by the cross-sectional area of the indentor. In those instances in which more than 1 peak was observed, this maximum was defined to be the first maximum value, followed by a drop in load of 10 % or more. The mean strength for each of the indentation sites was 26 to 67 MPa (Table I). One-way analysis of
variance indicated that there were significant differences in the strength of the bone among positions on the glenoid (F ¼ 14.0, P < .001). Similar results were observed for the elastic modulus, in which the mean value was 119 to 234 MPa, again, with a significant variation with position across the glenoid (F ¼ 7.6, P < .001). In general, there was a trend for the larger values of BV/TV to be found at the posterior margins of the glenoid, whereas the more porous
340
Mimar, Limb, and Hall
regions were observed near the anteroinferior edge. The differences in bone volume fraction across the glenoid were found to be significant (F ¼ 14.5, P <.001). The measured subchondral thickness was 0.88 mm to 1.15 mm, with thinner bone located inferiorly; again, analysis of variance demonstrated that the differences with position were significant. Multiple regression analysis, in which account was taken of the lack of independence of results taken from the same glenoid, indicated that both the bone volume fraction and the subchondral thickness have a significant effect on the elastic modulus. Similar results were observed for the relationship between strength and bone volume fraction and subchondral thickness. However, the R2 values of 0.21 for the elastic modulus and 0.31 for the strength analyses indicate that these linear models are poor and that other factors may play a significant role in determining the mechanical properties of the glenoid surface. DISCUSSION Regional differences were found in both the mechanical parameters and the morphologic measures, in keeping with previous studies.1,7 However, there were disparities in the absolute values of the parameters measured compared with those observed elsewhere. Frich et al7 reported mean ultimate strengths of approximately 35 MPa to more than 110 MPa, whereas the results reported here are roughly half these values. One possible reason for this discrepancy is the use of much younger specimens in the Frich et al11 cohort, which had a mean age of younger than 60 years. The results reported by Anglin et al1 were somewhat lower than those cited here or elsewhere and arise principally from the different methodology used in that multiple readings at different depths were taken, which included results from the weaker trabecular bone. Further disparities in both the elastic modulus and strength reported here and those accrued elsewhere can result from a number of other factors, including dissimilar specimens preparations, the removal of the subchondral bone,1 and significant variation in the experimental protocols and analyses.6 From this study, we believe that age and the use of female specimens are the most likely explanation for the thinner subchondral thickness we observed compared with past studies.5,7 To our knowledge, this is the first study to combine indentation and micro-CT analyses to investigate the relationship between the glenoid’s mechanical properties and bone morphology. The distribution of BV/ TV data is similar to that found for bone mineral density measurements by Lehtinen et al.14 However, the subchondral thickness measurements differ substantially from those previously reported7 and may result from differences in specimen preparation.
J Shoulder Elbow Surg March/April 2008
Both the local bone volume fraction and the subchondral thickness influenced the strength at the glenoid surface as well as the elastic modulus of the bone. The explained variance was small, however, and other factors not incorporated into this model may also have a significant effect on the strength and elastic modulus. These factors could include the degree of mineralization and architectural indices such as the anisotropy. In this particular set of experiments, embalmed cadaveric tissue was used. Anglin et al1 had previously suggested that this preservation had minimal impact on the quality of the results for this type of experiment, although it is difficult to compare the data reported here and those cited elsewhere to verify this assertion. If the embalming process were found to modify the mechanical properties of the tissue, then, provided this occurred in a relatively straightforward fashion, the relationships between strength or stiffness and the morphologic parameters would still likely hold true. Current work with fresh cadaveric tissue is ongoing to explore these issues further. Currently, the study has possibly identified that there is a band of bone across the glenoid that appears to provide superior mechanical properties for fixation according to the parameters we have identified. This runs from the base of the coracoid backward through the central glenoid to the posterior margin. This region should provide the best support for a glenoid component, although results will have to be verified on fresh bone. Previous studies by Frich6 have identified similar regional variation in bone properties, in which it was reported that the strongest bone lay posterior and superior, as well as anterior, to the bare area. Our study agrees with that of Frich in recognizing the poorer mechanical properties of bone at the inferior parts of the glenoid (Figure 5). It is notable that the anteroinferior region has the poorest structural and biomechanical properties. Interestingly, this lies within the region observed clinically that has a propensity for bone loss due to instability. This relationship merits further study. In summary, this study has identified a very useful combination of tools for studying this anatomic region that will facilitate development of more reliable glenoid fixation. We thank Dr Roger Soames for all of his contributions to this project.
REFERENCES
1. Anglin C, Tolhurst P, Wyss UP, Pichora DR. Glenoid cancellous bone strength and modulus. J Biomech 1999;32:1091-7. 2. Boyd SK, Muller R, Matyas JR, Wohl GR, Zernicke RF. Early morphometric and anisotropic change in periarticular cancellous bone in a model of experimental knee osteoarthritis quantified using microcomputed tomography. Clin Biomech 2000;15: 624-31.
J Shoulder Elbow Surg Volume 17, Number 2
3. Buchler P, Ramaniraka NA, Rakotomanana LR, Iannotti JP, Farron A. A finite element model of the shoulder: application to the comparison of normal and osteoarthritic joints. Clin Biomech 2002;17:630-9. 4. Churchill RS, Boorman RS, Fehringer EV, Matsen FA 3rd. Glenoid cementing may generate sufficient heat to endanger the surrounding bone. Clin Orthop Relat Res 2004:76-9. 5. Eckstein F, Milz S, Anetzberger H, Putz R. Thickness of the subchondral mineralised tissue zone (SMZ) in normal male and female and pathological human patellae. J Anat 1998;192:81-90. 6. Frich LH, Jensen NC, Odgaard A, Pedersen CM, Sojbjerg JO, Dalstra M. Bone strength and material properties of the glenoid. J Shoulder Elbow Surg 1997;6:97-104. 7. Frich LH, Odgaard A, Dalstra M. Glenoid bone architecture. J Shoulder Elbow Surg 1998;7:356-61. 8. Goldstein SA. The mechanical properties of trabecular bone: dependence on anatomic location and function. J Biomech 1987;20:1055-61. 9. Hasan SS, Leith JM, Campbell B, Kapil R, Smith KL, Matsen FA 3rd. Characteristics of unsatisfactory shoulder arthroplasties. J Shoulder Elbow Surg 2002;11:431-41. 10. Hvid I, Rasmussen O, Jensen NC, Nielsen S. Trabecular bone strength profiles at the ankle joint. Clin Orthop Relat Res 1985:306-12. 11. Ibarra C, Dines DM, McLaughlin JA. Glenoid replacement in total shoulder arthroplasty. Orthop Clin North Am 1998;29: 403-13.
Mimar, Limb, and Hall
341
12. Keaveny TM, Hayes WC. A 20-year perspective on the mechanical properties of trabecular bone. J Biomech Eng 1993; 115:534-42. 13. Lacroix D, Murphy LA, Prendergast PJ. Three-dimensional finite element analysis of glenoid replacement prostheses: a comparison of keeled and pegged anchorage systems. J Biomech Eng 2000; 122:430-6. 14. Lehtinen JT, Tingart MJ, Apreleva M, Warner JJ. Total, trabecular, and cortical bone mineral density in different regions of the glenoid. J Shoulder Elbow Surg 2004;13:344-8. 15. Linde F, Pongsoipetch B, Frich LH, Hvid I. Three-axial strain controlled testing applied to bone specimens from the proximal tibial epiphysis. J Biomech 1990;23:1167-72. 16. Linde F. Elastic and viscoelastic properties of trabecular bone by a compression testing approach. Dan Med Bull 1994;41: 119-38. 17. Mu¨ller R, Hahn M, Vogel M, Delling G, Ru¨egsegger P. Morphometric analysis of noninvasively assessed bone biopsies: comparison of high-resolution CT and histologic sections. Bone 1996;18: 215-20. 18. Terrier A, Buchler P, Farron A. Bone-cement interface of the glenoid component: stress analysis for varying cement thickness. Clin Biomech 2005;20:710-7. 19. Timoshenko SP, Goodier JN. Theory of elasticity. 3rd ed. New York, NY: McGraw-Hill; 1970. p. 408.