- Email: [email protected]
Acetabular Cup Stiffness and Implant Orientation Change Acetabular Loading Patterns
The Journal of Arthroplasty Vol. 28 No. 2 2013
Acetabular Cup Stiffness and Implant Orientation Change Acetabular Loading Patterns Scott R. Small, MS,y Michael E. Berend, MD,y Leah A. Howard, BS,* Didem Tunç, BS,* Christine A. Buckley, PhD,* and Merrill A. Ritter, MDy
Abstract: Acetabular cup orientation has been shown to influence dislocation, impingement, edge loading, contact stress, and polyethylene wear in total hip arthroplasty. Acetabular implant stiffness has been suggested as a factor in pelvic stress shielding and osseous integration. This study was designed to examine the combined effects of acetabular cup orientation and stiffness and on pelvic osseous loading. Four implant designs of varying stiffness were implanted into a composite hemipelvis in 35° or 50° of abduction. Specimens were dynamically loaded to simulate gait and pelvic strains were quantified with a grid of rosette strain gages and digital image correlation techniques. Changes in the joint reaction force orientation significantly altered mean acetabular bone strain values up to 67%. Increased cup abduction resulted in a 12% increase along the medial acetabular wall and an 18% decrease in strain in inferior lateral regions. Imbalanced loading distributions were observed with the stiffer components, resulting in higher, more variable, and localized surface strains. This study illustrates the effects of cup stiffness, gait, and implant orientation on loading distributions across the implanted pelvis. Keywords: stress shielding, acetabulum, pelvis, biomechanics, THA. © 2013 Elsevier Inc. All rights reserved.
Hip mechanics are a complex interaction of osseous and muscular forces across the joint. Joint reaction forces and articular contact mechanics have been previously studied [1-5]. Uncemented total hip arthroplasty (THA) components may uniquely alter the physiologic distribution of loads in the pelvis and femur and contribute to stress shielding and specific patterns of osseous remodeling. Femoral side effects have been studied clinically and in in vitro models; however, relatively few studies have quantified the distribution of loading in the pelvis after THA, particularly as it is influenced by component orientation, implant stiffness, and hip position during gait. Malalignment of the acetabular component has been correlated with clinical complications including increased dislocation rates [6-11], higher polymer wear, osteolysis, and hard-on-hard component noise generation [12-16]. Although these clinical correlations with cup alignment are well documented, insufficient inforFrom the *Rose-Hulman Institute of Technology, Department of Applied Biology and Biomedical Engineering, Terre Haute, Indiana; and yJoint Replacement Surgeons of Indiana Foundation, Inc, Mooresville, Indiana. Submitted November 17, 2011; accepted May 23, 2012. The Conflict of Interest statement associated with this article can be found at http://dx.doi.org/10.1016/j.arth.2012.05.026. Reprint requests: Michael E. Berend, MD, 1199 Hadley Road, Mooresville, IN 46158. © 2013 Elsevier Inc. All rights reserved. 0883-5403/2802-0027$36.00/0 http://dx.doi.org/10.1016/j.arth.2012.05.026
mation exists on the effect of acetabular cup orientation on the loading response and force distribution within the implanted pelvis. Although extensive computational and bench testing has been published on pelvic loading in the intact pelvis [17-21], data regarding pelvic loading after THA have been primarily generated from computational methods [22-26]. Fewer studies have used in vitro laboratory testing for the study of post-THA pelvic loading; specifically with respect to the influence of acetabular cup stiffness and inclination on loading patterns [27-29]. The purpose of this study is first, to establish a simplified model to quantify changes in (1) peak location and (2) magnitude of periacetabular strains. The second purpose is to quantify the alteration of strain distribution in the hemipelvis model due to changes in cup position, component stiffness for a series of acetabular component designs, and the angle of hip flexion. We hypothesize that specimens implanted with press-fit acetabular components will exhibit significantly altered loading with (1) stiffer cups decreasing pelvic strains and (2) increased abduction angles increasing strain in the roof and lateral margin of the acetabulum.
Methods Four groups of 8 composite left hemipelvis models (Model 3404, Large Left Fourth Generation Hemipelvis; Pacific Research Laboratories, Vashon, Wash) were fully
359
360 The Journal of Arthroplasty Vol. 28 No. 2 February 2013 Table. Acetabular Component Design and Geometry Implant Design
Substrate Material
Surface Material
Regenerex Ringloc Ranawat/Burstein M2a-38 M2a-Magnum
Ti-6Al-4V Ti-6Al-4V CoCr CoCr
Ti-6Al-4V Ti-6Al-4V Ti-6Al-4V Ti-6Al-4V
Substrate Thickness (mm)
Ingrowth Thickness (mm)
Overall Cup Thickness (mm)
Stiffness* (N/mm)
3.54 3.42 9.4 2.53/5.31 †
1.5 0.76 0.76 0.76
5.04 4.18 10.16 3.29/6.07 †
4598 2604 47244 5629
Articular Design Modular Modular Monoblock Monoblock
* Stiffness assessed at 1-kN rim compression by the manufacturer. † Thickness at rim/thickness at apex.
reamed with a 58-mm acetabular reamer and implanted with one 58-mm cup each of 4 designs: (1) RanawatBurstein (28-mm head size), (2) M 2a-38 (38-mm head size), (3) M 2a-Magnum (52-mm head size), or (4) Regenerex RingLoc (32-mm head size) acetabular components. The component features including materials and stiffness are listed in the Table. Acetabular reaming and component implantation were performed manually by a board-certified orthopedic surgeon. The acetabular components were implanted with standard instrumentation into 1 of 2 orientations: (1) 35° of abduction and anatomical anteversion, with the cups implanted flush with the lateral acetabular rim; (2) 50° of abduction and anatomical anteversion leaving 1 cm of exposed bone on the lateral acetabular rim. To quantify interspecimen variability in cup orientation after implantation, anterior-posterior radiographs were taken of each implanted hemipelvis specimen. Specimens were positioned with a soft foam mold on a patient x-ray table to align implanted hemipelvis specimens for cup abduction angle approximation (Fig. 1). A standard cup inclination measurement is not possible without a full pelvic radiograph; however, a horizontal reference line was approximated as perpendicular to the pubic symphysis. An additional linear measurement was taken
of the exposed reamed acetabulum from the edge of the acetabular shell to the acetabular rim at the lateral acetabular rim to provide a second method for orientation assessment. Strain measurements were recorded via a grid of eight 3-element rectangular rosette strain gages (KFG-3-120D17-11L3M2S; Kyowa, Tokyo, Japan) distributed across the surface of the implanted hemipelvis specimens as shown in Fig. 2. To ensure the repeatable placement of strain gages between each specimen, positioning and alignment guidelines were marked on the surface of each hemipelvis specimen before final gage fixation using a positioning mold rigidly affixed to the base plate of a coordinate measurement machine (BRM507; Mitutoyo America, Aurora, IL). This method of gage placement allowed for the repeatable placement of strain gages within ± 1 mm between each specimen. Data were recorded from each hemipelvis specimen over the course of 10 trials in each of the 2 loading positions for a total of 20 experimental trials per specimen using a strain gage data acquisition system (System 5000; Vishay MicroMeasurements, Raleigh, NC). In testing parallel to strain gage instrumented specimens, 4 hemipelvis specimens were implanted with 35° of abduction and anatomical anteversion and prepared
Fig. 1. Anterior-posterior radiographs of composite hemipelvis with (left) acetabular component implanted in 35° of abduction (strain gage wires visible) and (right) a component implanted in 50° of abduction.
Acetabular Cup Stiffness and Implant Orientation Small et al
361
Fig. 2. Strain gage positions on the (left) lateral and (right) medial aspects of the left hemipelvis model.
for full-field digital image correlation (DIC) strain analysis. Digital image correlation is a computational technique in which a series of high-resolution still images is captured throughout a loading cycle and analyzed for deformational on a specimen of interest. Deformational and strain measurements are calculated through the relative localized displacement of irregular black and white speckle patterns on the surface of a test specimen. The ARAMIS 5M (GOM Inc, Braunschweig, Germany) was used to develop a full-field assessment of strain response in the periacetabular region of the hemipelvis. Hemipelvis specimens were mounted on a customdesigned loading platform allowing for rigid fixation at the pubic symphysis and illiosacral joint (Fig. 3). A joint reaction force (JRF) of 1.5 kN was applied to each specimen through the appropriately matching femoral
head component for each acetabular cup design by means of a servohydraulic materials testing machine (Model 858; MTS Corporation, Eden Prairie, Minn). Two loading positions were used to simulate femoral motion during gait. The first JRF was defined as 1.5 kN oriented in 0° of femoral abduction and 0° of femoral flexion. The JRF 2 was modeled as 0° of femoral abduction and 15° of femoral flexion. Each applied JRF was ramped at a rate of 100 N/s until the final load was achieved in each loading position. Strain gage data were collected throughout the entire loading cycle, with final strain measurements and resulting final JRF held constant for 60 seconds after the target 1.5 kN was reached. In each trial, residual strains were induced during initial loading due to the inherent compliance within the test setup and were subtracted from the final strain readings. Data analysis was performed using a linear mixed effects regression model to identify the variables, which effect changes in strain at each strain gage location. Although rosette strain gages allow for quantification of principle strains and principle strain directions, the maximum shear strain will be the primary means of strain comparison in this report, as it represents a relative net sum of the strain state at each strain gage location.
Results
Fig. 3. Designated regions for rigid fixation at the illiosacral joint and pubic symphysis.
Acetabular component placement repeatability was statistically assessed using anterior-posterior radiograph angle measurements and linear measurements of exposed reamed acetabulum. Within the first implantation group, the mean acetabular cup abduction angle was measured to be 36.3° ± 1.3° (95% confidence interval), whereas the second implantation group was measured to have an actual component abduction angle of 51.1° ± 2.1° (95% confidence interval). There was no significant difference in abduction angles between each cup design within the 35° target group (P = .386) or the 50° target group (P = .427). The average exposed lateral reamed acetabulum was measured at 2.3 ± 0.6 mm in the 35° group and 9.5 ± 1.1 mm in the 50° group.
362 The Journal of Arthroplasty Vol. 28 No. 2 February 2013
Fig. 4. Least square means of maximum shear strain, dependent on JRF vector alignment, at each of the 8 strain gage locations. *Statistically significant difference in strain response generated by the varied force vectors (P b .05).
As a test of the sensitivity of this model, changes in strain based on loading vector orientation were first evaluated. As seen in Fig. 4, statistically significant differences in strain at 7 of the 8 strain gage positions were observed between JRF 1 and JRF 2 as averaged across all cup and implantation groups. The largest differences in strain were observed in periacetabular gage 4 and gage 5 at the ischium. At periacetabular
position 4, a 34% higher strain was induced when the JRF was oriented to replicate 15° of femoral flexion (P b .0001). Conversely, the same loading vector resulted in 67% lower local surface strain at strain gage 5, located on the ischium (P b .0001). A consistent increase between 3% and 13% were observed at the medial wall at gage locations 6 (P b .0001), 7 (P = .0298), and 8 (P b .0001) when the JRF modeling 15° of femoral flexion was applied.
Fig. 5. Least square means of maximum shear strain, dependent on acetabular component alignment, at each of the 8 strain gage locations. *Statistically significant difference in strain response generated by abduction (P b .05).
Acetabular Cup Stiffness and Implant Orientation Small et al
363
Fig. 6. Least square means of maximum shear strain at each of the 8 strain gage locations in the stiffest (M 2a-38) and least stiff (Ranawat/Burstein) acetabular cup designs. *Statistically significant differences in strain response generated by differences in cup design (P b .05).
Statistically significant differences in strain were identified in 5 of the 7 strain measurement positions when cups placed in 35° of abduction were compared with those placed in 50° of abduction (Fig. 5). A statistically significant difference was observed in 2 of the 4 primary lateral, periacetabular gage locations (gages 1-4). At
periacetabular gage location 3, an increased cup abduction angle resulted in 16% higher observed strain (P = .0030). Conversely, increased cup abduction induced 18% lower surface strains at strain gage location 4 (P = .0035). Cup inclination had the greatest effect at the ischial spine where strain measurements
Fig. 7. Least square means of maximum shear strain at each of the 8 strain gage locations statistically significant difference (P b .05) are indicated as: (a) greater than Regenerex, (b) greater than Ranawat/Burstein, (c) greater than M 2a-Magnum, and (d) greater than M 2a-38.
364 The Journal of Arthroplasty Vol. 28 No. 2 February 2013
Fig. 8. Digital image correlation measured periacetabular von Mises strain response in specimens with components in 35° of abduction and neutral anteversion, loaded to 1.5 kN in single-legged stance orientation. Regenerex RingLoc (A), RanawatBurstein (B), M 2a-Magnum (C), and M 2a-38 (D).
increased 108% at the ischium with an increased cup abduction angle (P b .0001). A primary end point of this study was to examine the effect of cup stiffness on the subsequent strain distribution within the pelvis. The M 2a-38 and the Ranawat/ Burstein modular cup designs represent the stiffest and least stiff of the cups tested in this study, respectfully. Fig. 6 shows the direct comparison between the strain measurements taken between the 2 designs, whereas least square means of strain response for all of the cup designs examined is shown in Fig. 7. At strain gage location 1 on the acetabular rim, the M 2a-38 design, the stiffest of the cups tested, exhibited significantly higher maximum shear strain than each of the other 3 designs (+ 31% vs M 2 a-Magnum, P = .0446, + 187% vs Ranawat/Burstein, P b .0001; + 104% vs Regenerex, P = .0002). Conversely, at this location, the Ranawat/ Burstein, the least stiff cup, demonstrated significantly lower strain responses than all other designs. The Ranawat/Burstein cup exhibited significantly higher strain than the Regenerex design at the acetabular rim at gage location 2, whereas the opposite was true further from the rim, at periacetabular strain gage location 3. Also more distant from the rim, both the Ranawat/ Burstein and M 2a-38 cups exhibited significantly higher strains than the Regenerex (gage 4). At the ischial spine (gage 5), cup stiffness played no statistically significant
role in measured strain response. The Regenerex cup generated significantly lower strain responses at the pubis (gage 6) and along the medial pelvic surface at strain gage location 8. However, significantly lower strains were measured at the medial pelvic surface in the two stiffest cup designs, the M 2a-Magnum and M 2a-38 cups at strain gage 7. As a parallel measurement system, DIC was used to yield a full-field observation of strain patterns in the periacetabular region of the pelvis as affected by acetabular component design and stiffness. As can be seen in the DIC von Mises strain maps shown in Fig. 8, the most even periacetabular, cortical loading distributions were observed in the most compliant Ranawat/ Burstein cup design with relatively low measured strains throughout the observation area and without any significant peak load hotspots indicating uneven loading distribution. While exhibiting a slightly higher overall strain response than the Ranawat/Burstein cup design, the Regenerex cup exhibited a similar trend in relatively smooth cortical loading distribution across the DIC measurement area and a general lack in localized strain hotspots. The two metal-on-metal cup designs tested, the M 2 a-38 and M 2 a-Magnum, exhibited higher overall strain responses, with mid-level to high localized strain at the acetabular rim and the greater sciatic notch.
Acetabular Cup Stiffness and Implant Orientation Small et al
Discussion
Understanding the impact of acetabular component alignment and component design and resulting implant stiffness on loading distribution in the pelvis is necessary for the long-term preservation of primary THA procedures. It was hypothesized that stiffer components would lead to increased periacetabular strains, with increased inclination altering strain in the roof and lateral margin. The results did not show uniform trends across all measurement locations as hypothesized; however, less stiff components were observed to generate more uniform periacetabular strain patterns. Counter to the expected result, component inclination generated minimal significant changes in strain response in the acetabular roof and lateral margin. Stress shielding in the acetabulum has been reported in the clinical literature and may lead to diminished bone stock or failure of cementless fixation or perhaps play a role in the osteolytic processes through wear debris into the underlying bone [30-32]. As they relate to acetabular loading distribution and stress shielding, three primary parameters were examined in this study: JRF vector, acetabular component abduction, and acetabular component stiffness. The orientation of the JRF vector served as a sensitivity test for subsequent testing. Joint reaction force was held constant at 1500 N, a level consistent with 238% body weight of a 70-kg patient, a level described by Bergmann et al [3] and Pedersen et al [5] as a typical contact force when walking at 4 km/h. Directional vectors were compared between a single-legged stance orientation as defined by McLeish [1] and a relative shift reflecting 15° for femoral flexion. Strain measurements for this parameter reflected a high sensitivity to change in JRF 7 of 8 measurement regions exhibiting statistically significant change in strain following a shift in reaction force orientation. In the examination of acetabular component alignment, the greatest significant changes in surface strains were observed in superior and posterior measurement regions as well as in the ischial body. In a specimenspecific finite element model, Zhang et al [33] reported highest pelvic strains in the periacetabular regions near the superior aspect of the acetabular rim. Interestingly, high cup abduction was not a significant contributor to changes in strain in regions closest to the acetabular rim, which may be in closer contact to the implanted component and potentially have greater impact than more medial and central regions of the pelvis. These areas of high strain would indicate a potential focal point for studies examining the resulting changes from stress shielding. A lack in periacetabular sensitivity in this region to changes in acetabular abduction between 35° and 50°, while potentially a side effect of this specific model design, could be a clinically important factor when relating component positioning with long-term pelvic stress shielding.
365
The primary mechanical contributor to stress shielding in vivo is the dramatic difference in modulus of elasticity between the native bone tissue and the much stiffer alloys used in current prosthetic designs. In this study, four acetabular component designs with wide ranging global stiffness were selected to examine the varied effects of component design on pelvic strain distribution. Titanium cups with modular polyethylene inserts, the most pliable of tested constructs, generated more evenly distributed strain responses in most observation points, whereas the thick-walled metal-on-metal cup exhibited nearly twice the strain response in the periacetabular regions posteriorly and toward the greater sciatic notch. Digital image correlation data exhibit more evenly distributed periacetabular strains, with lower variation between gage locations in the less stiff porous titanium construct, with increased periacetabular variation in the stiffer CoCrMo metal-on-metal cups. Similar findings were observed in a comparison of CoCr, polyester ester ketone (PEEK), and polyethylene cups in a recent DIC study [27]. These results correlate well with a prior finite element study, in which Manley et al [34] performed an acetabular stress analysis on a specimen-specific model from a young female cadaveric specimen. In this model, they found significantly more evenly distributed stresses in the most compliant cups, yet nonphysiologic loading distribution leading to the potential for bone resorption with all hemispherical shell materials tested. The alteration of loading based on stiffer acetabular constructs can be linked to the press-fit implantation methods common to uncemented THA. A construct with elastic modulus closer to that of the native bone should allow for more natural load transfer at the boneimplant interface due to the matching material properties. Conversely, a vastly higher modulus in stiff acetabular constructs should generate more “pushback” at the rim, generating focal points of high stress, while potentially shunting load away from other supporting regions, generating nonphysiologic remodeling and reduced overall bone quality [35]. Most studies addressing stress shielding and pelvic loading have focused on the measurement of bone mineral density (BMD) rather than the characterization of pelvic loading, with varied results. Periacetabular bone loss due to stress shielding has been observed up to 34% at 1 year after THA and is correlated with nonphysiologic distribution of periacetabular loads postoperatively [32,36,37]. In a recent randomized, prospective study, Meneghini et al [35] reported change in BMD at a mean follow-up of 7.7 years in the anterosuperior and posterosuperior periacetabular regions between solid titanium and porous tantalum acetabular components. The authors observed significant stress shielding in 80% of the measured areas when implanted with the stiffer acetabular component. This long-term decay of BMD related to stiffer acetabular
366 The Journal of Arthroplasty Vol. 28 No. 2 February 2013 constructs correlates well with the current study and validates the need for increased understanding of localized changes in bony load distribution and its resultant effect on long-term stress shielding around the implant. There are study limitations in the design of this model. Primarily, this study uses rigid fixation of a hemipelvis at the pubic symphysis and illiosacral joint. This fixation method is driven by the utilization of a composite bone specimen without physiologic ligamentous attachments or muscular interactions. The utilization of rigid fixation without muscular and ligamentous characterization has been documented to generate an overall stiffer strain response than the physiologic condition and more complex setup arrangements [21]. Consequently, this model is prone to exhibit more pronounced localized strain “hotspots” than might be physiologically accurate. Digital image correlation was used as a supplemental data acquisition technique to fill the voids in data between individual strain gage measurement locations. Statistically analyzed data in this study were exclusively gathered through standard strain gage techniques, and as a result, only reflects a small subset of the actual strain response across the entirety of the analog specimen. Results from DIC analysis indicate substantial variation in strain distribution across the pelvis in most loading scenarios. This substantial variation in strain response, even in well-distributed loading from compliant device designs, is reflected in strain gage data through high variability at each strain gage location. In addition, nonimplanted initial strain measurements were not taken, as the authors felt that a lack of a physiologic articular surface at the acetabulum prohibited sufficiently accurate unimplanted loading characteristics. The lack of baseline measurements in the unimplanted hemipelvis may limit the study of actual stress shielding characteristics. Likewise, as an analog model, the longterm biological remodeling response of the living pelvis cannot be replicated. Nevertheless, the tightly controlled nature of this setup, using a repeatable composite hemipelvis setup, allows for the direct comparison of implant alignment, JRF orientation, and acetabular component design on the change of pelvic strain response in a consistent specimen model. The primary focus of this study was to establish a simplified pelvic loading model allowing quantification and anatomical mapping of periacetabular strains, while simultaneously enabling the quantification of the change in cortical loading as a result of varied surgical and prosthetic parameters. The experimental model in this study used composite hemipelvis models from a family of analog test specimens validated against intact cadaveric bone tissues [38,39] and responded in a repeatable manner when tested for sensitivity through varying JRF vectors. Although this model does not offer the robust characteristics of an advanced computational
model nor the physiologic accuracy of a whole-pelvis cadaveric model, it does provide a uniform and consistent test medium for the preliminary comparison of relative changes in strain response due to alterations in prosthetic design and alignment. In summary, this study generated a simplified mechanical model, sensitive to loading, orientation, and component stiffness, to effectively compare the relationship between surgical and prosthetic factors and their role in cortical strain distribution in the total hip arthroplasty pelvic model. Strain gage and DIC data analysis techniques identified global implant stiffness as the greatest contributor to strain distribution across the periacetabular regions of the pelvis at the time of implantation, with stiffer acetabular components generating elevated strains at the acetabular rim in stiffer components and relatively more even distribution in the more compliant cup designs. Strain response due to orientation of the applied JRF and component abduction were less apparent near the acetabular rim, with most significant response changes occurring nearer to the body of the ilium as well as the medial wall. As the market for primary and revision THA continues to expand, a full understanding of the surgical and designspecific implications on pelvic loading, stress shielding, and the preservation of bone stock is imperative. Further study in the area of pelvic loading and stress shielding is needed to support the long-term success of cementless THA.
References 1. McLeish RD. Abduction forces in the one-legged stance. J Biomechanics 1970;3:191. 2. Hodge WA, Carlson KL, Fijan RS, et al. Contact pressures from an instrumented hip endoprosthesis. J Bone Joint Surg 1989;71:1378. 3. Bergmann G, Graichen F, Rohlmann A. Hip joint loading during walking and running, measured in two patients. J Biomechanics 1993;26:969. 4. Olson SA, Bay BK, Hamel A. Biomechanics of the hip joint and effects of fracture of the acetabulum. Clin Orthop Relat Res 1997;338:92. 5. Pedersen DR, Brand RA, Davy DT. Pelvic muscle and acetabular contact forces during gait. J Biomechanics 1997;30:959. 6. Charnley J. Total hip replacement by low-friction arthroplasty. Clin Orthop Relat Res 1970;72:7. 7. Muller ME. Total hip prosthesis. Clin Orthop Relat Res 1970;72:46. 8. Coventry MB, Beckenbaugh RD, Nolan DR, et al. 2,012 total hip arthroplasties. A study of postoperative course and early complications. J Bone Joint Surg Am 1974;56:273. 9. Harris WH. Advances in surgical technique for total hip replacement: without and with osteotomy of the greater trochanter. Clin Orthop Relat Res 1980;146:188. 10. Harkess JW. Arthroplasty of hip: surgical approaches and techniques. In: Canale SD, editor. Campbell's Operative Orthopaedics. 10th ed. St Louis, Mo: Mosby; 2003; p. 348.
Acetabular Cup Stiffness and Implant Orientation Small et al 11. Lewinnek GE, Lewis JL, Tarr R, et al. Dislocations after total hip replacement arthroplasties. J Bone Joint Surg Am 1978;60:217. 12. D'Lima DD, Urquhart AG, Buehler KO, et al. The effect of the orientation of the acetabular and femoral components on the range of motion of the hip at different head-neck ratios. J Bone Joint Surg Am 2000;82:315. 13. Dorr LD, Wan Z. Causes of and treatment protocol for instability of total hip replacement. Clin Orthop Relat Res 1998;355:144. 14. Paterno SA, Lachiewicz PF, Kelley SS. The influence of 490 patient-related factors and the position of the acetabular component on the Q9 rate of dislocation. J Bone Joint Surg Am 1997;78:163. 15. Sexton SA, Yeung E, Jackson MP, et al. The role of patient factors and implant position in squeaking of ceramic-onceramic total hip replacements. J Bone Joint Surg Br 2011; 93-B:439. 16. Walter WL, O'toole GC, Walter WK, et al. Squeaking in ceramic-on-ceramic hips: the importance of acetabular component orientation. J Arthroplasty 2007;22:496. 17. Anderson AE, Peters CL, Tuttle BD, et al. Subject-specific finite element model of the pelvis: development, validation and sensitivity studies. J Biomech Eng 2005;127:364. 18. Dalstra M, Huiskes R, van Erning L. Development and validation of a three-dimensional finite element model of the pelvic bone. Transactions of the ASME. J Biomech Eng 1995;117:272. 19. Dalstra M, Huiskes R. Load transfer across the pelvic bone. J Biomech 1995;28:715. 20. Oonishi H, Isha H, Hasegawa T. Mechanical analysis of the human pelvis and its application to the artificial hip joint— by means of the three dimensional finite element method. J Biomech 1983;16:427. 21. Phillips ATM, Pankaj P, Howie CR, et al. Finite element modeling of the pelvis: inclusion of muscular and ligamentous boundary conditions. Med Eng Phys 2007;29:739. 22. Thompson MS, Northmore-Ball MD, Tanner KE. Effects of acetabular resurfacing component material and fixation on the strain distribution in the pelvis. Proc Inst Mech Eng H 2002;216:237. 23. Vasu R, Carter DR, Harris WH. Stress distributions in the acetabular region—I. Before and after joint replacement. J Biomech 1982;15:155. 24. Carter DR, Vasu R, Harris WH. Stress distributions in the acetabular region—II. Effects of cement thickness and metal backing of the total hip acetabular component. J Biomech 1982;15:165.
367
25. Levenston ME, Beaupre GS, Schurman DJ, et al. Computer simulations of stress-related bone remodeling around noncemented acetabular components. J Arthroplasty 1993;8:595. 26. Huiskes R. Finite element analysis of acetabular reconstruction: noncemented threaded cups. Acta Orthop 1987; 58:620. 27. Dickinson AS, Taylor AC, Browne M. The influence of acetabular cup material on pelvis cortex surface strains, measured using digital image correlation. J Biomech 2012; 45:719. 28. Lionberger D, Walker PS, Granholm J. Effects of prosthetic acetabular replacement on strains in the pelvis. J Orthopaedic Research 1985;3:372. 29. Slemon D, O'Donnell P, Little E. The use of photoelastic coatings to investigate the pelvic strain variations resulting from the introduction of cementless metal backed acetabular cups. Strain 1991;27:143. 30. Mjoberg B. The theory of early loosening of implants. Orthopaedics 1997;20:1169. 31. Schmidt R, Muller L, Kress A, et al. A computed tomography assessment of femoral and acetabular bone changes after total hip arthroplasty. Int Orthop 2002; 26:299. 32. Wright JM, Pellicci PM, Salvati EA, et al. Bone density adjacent to press-fit acetabular components. A prospective analysis with quantitative computed tomography. J Bone Joint Surg 2001;83-A:529. 33. Zhang Q, Wang J, Lupton C, et al. A subject-specific pelvic bone model and its application to cemented acetabular replacements. J Biomechanics 2010;43:2722. 34. Manley MT, Ong KL, Kurtz SM. The potential for bone loss in acetabular structures following THA. Clin Orthop Relat Res 2006;453:246. 35. Meneghini RM, Ford KS, McCollough CH, et al. Bone remodeling around porous metal cementless acetabular components. J Arthroplasty 2010;25:741. 36. Pitto RP, Bhargava A, Pandit S, et al. Retroacetabular stress-shielding in THA. Clin Orthop Relat Res 2008; 466:353. 37. Stepniewski AS, Egawa H, Sychterz-Terefenko C, et al. Periacetabular bone density after total hip arthroplasty: a postmortem analysis. J Arthroplasty 2008;23:593. 38. Heiner AD. Structural properties of fourth-generation composite femurs and tibias. J Biomech 2008;41:3282. 39. Gardner MP, Chong ACM, Pollock AG, et al. Mechanical evaluation of large-size fourth-generation composite femur and tibia models. Ann Biomed Eng 2010;38:618.
Acetabular Cup Stiffness and Implant Orientation Change Acetabular Loading Patterns Scott R. Small, MS,y Michael E. Berend, MD,y Leah A. Howard, BS,* Didem Tunç, BS,* Christine A. Buckley, PhD,* and Merrill A. Ritter, MDy
Abstract: Acetabular cup orientation has been shown to influence dislocation, impingement, edge loading, contact stress, and polyethylene wear in total hip arthroplasty. Acetabular implant stiffness has been suggested as a factor in pelvic stress shielding and osseous integration. This study was designed to examine the combined effects of acetabular cup orientation and stiffness and on pelvic osseous loading. Four implant designs of varying stiffness were implanted into a composite hemipelvis in 35° or 50° of abduction. Specimens were dynamically loaded to simulate gait and pelvic strains were quantified with a grid of rosette strain gages and digital image correlation techniques. Changes in the joint reaction force orientation significantly altered mean acetabular bone strain values up to 67%. Increased cup abduction resulted in a 12% increase along the medial acetabular wall and an 18% decrease in strain in inferior lateral regions. Imbalanced loading distributions were observed with the stiffer components, resulting in higher, more variable, and localized surface strains. This study illustrates the effects of cup stiffness, gait, and implant orientation on loading distributions across the implanted pelvis. Keywords: stress shielding, acetabulum, pelvis, biomechanics, THA. © 2013 Elsevier Inc. All rights reserved.
Hip mechanics are a complex interaction of osseous and muscular forces across the joint. Joint reaction forces and articular contact mechanics have been previously studied [1-5]. Uncemented total hip arthroplasty (THA) components may uniquely alter the physiologic distribution of loads in the pelvis and femur and contribute to stress shielding and specific patterns of osseous remodeling. Femoral side effects have been studied clinically and in in vitro models; however, relatively few studies have quantified the distribution of loading in the pelvis after THA, particularly as it is influenced by component orientation, implant stiffness, and hip position during gait. Malalignment of the acetabular component has been correlated with clinical complications including increased dislocation rates [6-11], higher polymer wear, osteolysis, and hard-on-hard component noise generation [12-16]. Although these clinical correlations with cup alignment are well documented, insufficient inforFrom the *Rose-Hulman Institute of Technology, Department of Applied Biology and Biomedical Engineering, Terre Haute, Indiana; and yJoint Replacement Surgeons of Indiana Foundation, Inc, Mooresville, Indiana. Submitted November 17, 2011; accepted May 23, 2012. The Conflict of Interest statement associated with this article can be found at http://dx.doi.org/10.1016/j.arth.2012.05.026. Reprint requests: Michael E. Berend, MD, 1199 Hadley Road, Mooresville, IN 46158. © 2013 Elsevier Inc. All rights reserved. 0883-5403/2802-0027$36.00/0 http://dx.doi.org/10.1016/j.arth.2012.05.026
mation exists on the effect of acetabular cup orientation on the loading response and force distribution within the implanted pelvis. Although extensive computational and bench testing has been published on pelvic loading in the intact pelvis [17-21], data regarding pelvic loading after THA have been primarily generated from computational methods [22-26]. Fewer studies have used in vitro laboratory testing for the study of post-THA pelvic loading; specifically with respect to the influence of acetabular cup stiffness and inclination on loading patterns [27-29]. The purpose of this study is first, to establish a simplified model to quantify changes in (1) peak location and (2) magnitude of periacetabular strains. The second purpose is to quantify the alteration of strain distribution in the hemipelvis model due to changes in cup position, component stiffness for a series of acetabular component designs, and the angle of hip flexion. We hypothesize that specimens implanted with press-fit acetabular components will exhibit significantly altered loading with (1) stiffer cups decreasing pelvic strains and (2) increased abduction angles increasing strain in the roof and lateral margin of the acetabulum.
Methods Four groups of 8 composite left hemipelvis models (Model 3404, Large Left Fourth Generation Hemipelvis; Pacific Research Laboratories, Vashon, Wash) were fully
359
360 The Journal of Arthroplasty Vol. 28 No. 2 February 2013 Table. Acetabular Component Design and Geometry Implant Design
Substrate Material
Surface Material
Regenerex Ringloc Ranawat/Burstein M2a-38 M2a-Magnum
Ti-6Al-4V Ti-6Al-4V CoCr CoCr
Ti-6Al-4V Ti-6Al-4V Ti-6Al-4V Ti-6Al-4V
Substrate Thickness (mm)
Ingrowth Thickness (mm)
Overall Cup Thickness (mm)
Stiffness* (N/mm)
3.54 3.42 9.4 2.53/5.31 †
1.5 0.76 0.76 0.76
5.04 4.18 10.16 3.29/6.07 †
4598 2604 47244 5629
Articular Design Modular Modular Monoblock Monoblock
* Stiffness assessed at 1-kN rim compression by the manufacturer. † Thickness at rim/thickness at apex.
reamed with a 58-mm acetabular reamer and implanted with one 58-mm cup each of 4 designs: (1) RanawatBurstein (28-mm head size), (2) M 2a-38 (38-mm head size), (3) M 2a-Magnum (52-mm head size), or (4) Regenerex RingLoc (32-mm head size) acetabular components. The component features including materials and stiffness are listed in the Table. Acetabular reaming and component implantation were performed manually by a board-certified orthopedic surgeon. The acetabular components were implanted with standard instrumentation into 1 of 2 orientations: (1) 35° of abduction and anatomical anteversion, with the cups implanted flush with the lateral acetabular rim; (2) 50° of abduction and anatomical anteversion leaving 1 cm of exposed bone on the lateral acetabular rim. To quantify interspecimen variability in cup orientation after implantation, anterior-posterior radiographs were taken of each implanted hemipelvis specimen. Specimens were positioned with a soft foam mold on a patient x-ray table to align implanted hemipelvis specimens for cup abduction angle approximation (Fig. 1). A standard cup inclination measurement is not possible without a full pelvic radiograph; however, a horizontal reference line was approximated as perpendicular to the pubic symphysis. An additional linear measurement was taken
of the exposed reamed acetabulum from the edge of the acetabular shell to the acetabular rim at the lateral acetabular rim to provide a second method for orientation assessment. Strain measurements were recorded via a grid of eight 3-element rectangular rosette strain gages (KFG-3-120D17-11L3M2S; Kyowa, Tokyo, Japan) distributed across the surface of the implanted hemipelvis specimens as shown in Fig. 2. To ensure the repeatable placement of strain gages between each specimen, positioning and alignment guidelines were marked on the surface of each hemipelvis specimen before final gage fixation using a positioning mold rigidly affixed to the base plate of a coordinate measurement machine (BRM507; Mitutoyo America, Aurora, IL). This method of gage placement allowed for the repeatable placement of strain gages within ± 1 mm between each specimen. Data were recorded from each hemipelvis specimen over the course of 10 trials in each of the 2 loading positions for a total of 20 experimental trials per specimen using a strain gage data acquisition system (System 5000; Vishay MicroMeasurements, Raleigh, NC). In testing parallel to strain gage instrumented specimens, 4 hemipelvis specimens were implanted with 35° of abduction and anatomical anteversion and prepared
Fig. 1. Anterior-posterior radiographs of composite hemipelvis with (left) acetabular component implanted in 35° of abduction (strain gage wires visible) and (right) a component implanted in 50° of abduction.
Acetabular Cup Stiffness and Implant Orientation Small et al
361
Fig. 2. Strain gage positions on the (left) lateral and (right) medial aspects of the left hemipelvis model.
for full-field digital image correlation (DIC) strain analysis. Digital image correlation is a computational technique in which a series of high-resolution still images is captured throughout a loading cycle and analyzed for deformational on a specimen of interest. Deformational and strain measurements are calculated through the relative localized displacement of irregular black and white speckle patterns on the surface of a test specimen. The ARAMIS 5M (GOM Inc, Braunschweig, Germany) was used to develop a full-field assessment of strain response in the periacetabular region of the hemipelvis. Hemipelvis specimens were mounted on a customdesigned loading platform allowing for rigid fixation at the pubic symphysis and illiosacral joint (Fig. 3). A joint reaction force (JRF) of 1.5 kN was applied to each specimen through the appropriately matching femoral
head component for each acetabular cup design by means of a servohydraulic materials testing machine (Model 858; MTS Corporation, Eden Prairie, Minn). Two loading positions were used to simulate femoral motion during gait. The first JRF was defined as 1.5 kN oriented in 0° of femoral abduction and 0° of femoral flexion. The JRF 2 was modeled as 0° of femoral abduction and 15° of femoral flexion. Each applied JRF was ramped at a rate of 100 N/s until the final load was achieved in each loading position. Strain gage data were collected throughout the entire loading cycle, with final strain measurements and resulting final JRF held constant for 60 seconds after the target 1.5 kN was reached. In each trial, residual strains were induced during initial loading due to the inherent compliance within the test setup and were subtracted from the final strain readings. Data analysis was performed using a linear mixed effects regression model to identify the variables, which effect changes in strain at each strain gage location. Although rosette strain gages allow for quantification of principle strains and principle strain directions, the maximum shear strain will be the primary means of strain comparison in this report, as it represents a relative net sum of the strain state at each strain gage location.
Results
Fig. 3. Designated regions for rigid fixation at the illiosacral joint and pubic symphysis.
Acetabular component placement repeatability was statistically assessed using anterior-posterior radiograph angle measurements and linear measurements of exposed reamed acetabulum. Within the first implantation group, the mean acetabular cup abduction angle was measured to be 36.3° ± 1.3° (95% confidence interval), whereas the second implantation group was measured to have an actual component abduction angle of 51.1° ± 2.1° (95% confidence interval). There was no significant difference in abduction angles between each cup design within the 35° target group (P = .386) or the 50° target group (P = .427). The average exposed lateral reamed acetabulum was measured at 2.3 ± 0.6 mm in the 35° group and 9.5 ± 1.1 mm in the 50° group.
362 The Journal of Arthroplasty Vol. 28 No. 2 February 2013
Fig. 4. Least square means of maximum shear strain, dependent on JRF vector alignment, at each of the 8 strain gage locations. *Statistically significant difference in strain response generated by the varied force vectors (P b .05).
As a test of the sensitivity of this model, changes in strain based on loading vector orientation were first evaluated. As seen in Fig. 4, statistically significant differences in strain at 7 of the 8 strain gage positions were observed between JRF 1 and JRF 2 as averaged across all cup and implantation groups. The largest differences in strain were observed in periacetabular gage 4 and gage 5 at the ischium. At periacetabular
position 4, a 34% higher strain was induced when the JRF was oriented to replicate 15° of femoral flexion (P b .0001). Conversely, the same loading vector resulted in 67% lower local surface strain at strain gage 5, located on the ischium (P b .0001). A consistent increase between 3% and 13% were observed at the medial wall at gage locations 6 (P b .0001), 7 (P = .0298), and 8 (P b .0001) when the JRF modeling 15° of femoral flexion was applied.
Fig. 5. Least square means of maximum shear strain, dependent on acetabular component alignment, at each of the 8 strain gage locations. *Statistically significant difference in strain response generated by abduction (P b .05).
Acetabular Cup Stiffness and Implant Orientation Small et al
363
Fig. 6. Least square means of maximum shear strain at each of the 8 strain gage locations in the stiffest (M 2a-38) and least stiff (Ranawat/Burstein) acetabular cup designs. *Statistically significant differences in strain response generated by differences in cup design (P b .05).
Statistically significant differences in strain were identified in 5 of the 7 strain measurement positions when cups placed in 35° of abduction were compared with those placed in 50° of abduction (Fig. 5). A statistically significant difference was observed in 2 of the 4 primary lateral, periacetabular gage locations (gages 1-4). At
periacetabular gage location 3, an increased cup abduction angle resulted in 16% higher observed strain (P = .0030). Conversely, increased cup abduction induced 18% lower surface strains at strain gage location 4 (P = .0035). Cup inclination had the greatest effect at the ischial spine where strain measurements
Fig. 7. Least square means of maximum shear strain at each of the 8 strain gage locations statistically significant difference (P b .05) are indicated as: (a) greater than Regenerex, (b) greater than Ranawat/Burstein, (c) greater than M 2a-Magnum, and (d) greater than M 2a-38.
364 The Journal of Arthroplasty Vol. 28 No. 2 February 2013
Fig. 8. Digital image correlation measured periacetabular von Mises strain response in specimens with components in 35° of abduction and neutral anteversion, loaded to 1.5 kN in single-legged stance orientation. Regenerex RingLoc (A), RanawatBurstein (B), M 2a-Magnum (C), and M 2a-38 (D).
increased 108% at the ischium with an increased cup abduction angle (P b .0001). A primary end point of this study was to examine the effect of cup stiffness on the subsequent strain distribution within the pelvis. The M 2a-38 and the Ranawat/ Burstein modular cup designs represent the stiffest and least stiff of the cups tested in this study, respectfully. Fig. 6 shows the direct comparison between the strain measurements taken between the 2 designs, whereas least square means of strain response for all of the cup designs examined is shown in Fig. 7. At strain gage location 1 on the acetabular rim, the M 2a-38 design, the stiffest of the cups tested, exhibited significantly higher maximum shear strain than each of the other 3 designs (+ 31% vs M 2 a-Magnum, P = .0446, + 187% vs Ranawat/Burstein, P b .0001; + 104% vs Regenerex, P = .0002). Conversely, at this location, the Ranawat/ Burstein, the least stiff cup, demonstrated significantly lower strain responses than all other designs. The Ranawat/Burstein cup exhibited significantly higher strain than the Regenerex design at the acetabular rim at gage location 2, whereas the opposite was true further from the rim, at periacetabular strain gage location 3. Also more distant from the rim, both the Ranawat/ Burstein and M 2a-38 cups exhibited significantly higher strains than the Regenerex (gage 4). At the ischial spine (gage 5), cup stiffness played no statistically significant
role in measured strain response. The Regenerex cup generated significantly lower strain responses at the pubis (gage 6) and along the medial pelvic surface at strain gage location 8. However, significantly lower strains were measured at the medial pelvic surface in the two stiffest cup designs, the M 2a-Magnum and M 2a-38 cups at strain gage 7. As a parallel measurement system, DIC was used to yield a full-field observation of strain patterns in the periacetabular region of the pelvis as affected by acetabular component design and stiffness. As can be seen in the DIC von Mises strain maps shown in Fig. 8, the most even periacetabular, cortical loading distributions were observed in the most compliant Ranawat/ Burstein cup design with relatively low measured strains throughout the observation area and without any significant peak load hotspots indicating uneven loading distribution. While exhibiting a slightly higher overall strain response than the Ranawat/Burstein cup design, the Regenerex cup exhibited a similar trend in relatively smooth cortical loading distribution across the DIC measurement area and a general lack in localized strain hotspots. The two metal-on-metal cup designs tested, the M 2 a-38 and M 2 a-Magnum, exhibited higher overall strain responses, with mid-level to high localized strain at the acetabular rim and the greater sciatic notch.
Acetabular Cup Stiffness and Implant Orientation Small et al
Discussion
Understanding the impact of acetabular component alignment and component design and resulting implant stiffness on loading distribution in the pelvis is necessary for the long-term preservation of primary THA procedures. It was hypothesized that stiffer components would lead to increased periacetabular strains, with increased inclination altering strain in the roof and lateral margin. The results did not show uniform trends across all measurement locations as hypothesized; however, less stiff components were observed to generate more uniform periacetabular strain patterns. Counter to the expected result, component inclination generated minimal significant changes in strain response in the acetabular roof and lateral margin. Stress shielding in the acetabulum has been reported in the clinical literature and may lead to diminished bone stock or failure of cementless fixation or perhaps play a role in the osteolytic processes through wear debris into the underlying bone [30-32]. As they relate to acetabular loading distribution and stress shielding, three primary parameters were examined in this study: JRF vector, acetabular component abduction, and acetabular component stiffness. The orientation of the JRF vector served as a sensitivity test for subsequent testing. Joint reaction force was held constant at 1500 N, a level consistent with 238% body weight of a 70-kg patient, a level described by Bergmann et al [3] and Pedersen et al [5] as a typical contact force when walking at 4 km/h. Directional vectors were compared between a single-legged stance orientation as defined by McLeish [1] and a relative shift reflecting 15° for femoral flexion. Strain measurements for this parameter reflected a high sensitivity to change in JRF 7 of 8 measurement regions exhibiting statistically significant change in strain following a shift in reaction force orientation. In the examination of acetabular component alignment, the greatest significant changes in surface strains were observed in superior and posterior measurement regions as well as in the ischial body. In a specimenspecific finite element model, Zhang et al [33] reported highest pelvic strains in the periacetabular regions near the superior aspect of the acetabular rim. Interestingly, high cup abduction was not a significant contributor to changes in strain in regions closest to the acetabular rim, which may be in closer contact to the implanted component and potentially have greater impact than more medial and central regions of the pelvis. These areas of high strain would indicate a potential focal point for studies examining the resulting changes from stress shielding. A lack in periacetabular sensitivity in this region to changes in acetabular abduction between 35° and 50°, while potentially a side effect of this specific model design, could be a clinically important factor when relating component positioning with long-term pelvic stress shielding.
365
The primary mechanical contributor to stress shielding in vivo is the dramatic difference in modulus of elasticity between the native bone tissue and the much stiffer alloys used in current prosthetic designs. In this study, four acetabular component designs with wide ranging global stiffness were selected to examine the varied effects of component design on pelvic strain distribution. Titanium cups with modular polyethylene inserts, the most pliable of tested constructs, generated more evenly distributed strain responses in most observation points, whereas the thick-walled metal-on-metal cup exhibited nearly twice the strain response in the periacetabular regions posteriorly and toward the greater sciatic notch. Digital image correlation data exhibit more evenly distributed periacetabular strains, with lower variation between gage locations in the less stiff porous titanium construct, with increased periacetabular variation in the stiffer CoCrMo metal-on-metal cups. Similar findings were observed in a comparison of CoCr, polyester ester ketone (PEEK), and polyethylene cups in a recent DIC study [27]. These results correlate well with a prior finite element study, in which Manley et al [34] performed an acetabular stress analysis on a specimen-specific model from a young female cadaveric specimen. In this model, they found significantly more evenly distributed stresses in the most compliant cups, yet nonphysiologic loading distribution leading to the potential for bone resorption with all hemispherical shell materials tested. The alteration of loading based on stiffer acetabular constructs can be linked to the press-fit implantation methods common to uncemented THA. A construct with elastic modulus closer to that of the native bone should allow for more natural load transfer at the boneimplant interface due to the matching material properties. Conversely, a vastly higher modulus in stiff acetabular constructs should generate more “pushback” at the rim, generating focal points of high stress, while potentially shunting load away from other supporting regions, generating nonphysiologic remodeling and reduced overall bone quality [35]. Most studies addressing stress shielding and pelvic loading have focused on the measurement of bone mineral density (BMD) rather than the characterization of pelvic loading, with varied results. Periacetabular bone loss due to stress shielding has been observed up to 34% at 1 year after THA and is correlated with nonphysiologic distribution of periacetabular loads postoperatively [32,36,37]. In a recent randomized, prospective study, Meneghini et al [35] reported change in BMD at a mean follow-up of 7.7 years in the anterosuperior and posterosuperior periacetabular regions between solid titanium and porous tantalum acetabular components. The authors observed significant stress shielding in 80% of the measured areas when implanted with the stiffer acetabular component. This long-term decay of BMD related to stiffer acetabular
366 The Journal of Arthroplasty Vol. 28 No. 2 February 2013 constructs correlates well with the current study and validates the need for increased understanding of localized changes in bony load distribution and its resultant effect on long-term stress shielding around the implant. There are study limitations in the design of this model. Primarily, this study uses rigid fixation of a hemipelvis at the pubic symphysis and illiosacral joint. This fixation method is driven by the utilization of a composite bone specimen without physiologic ligamentous attachments or muscular interactions. The utilization of rigid fixation without muscular and ligamentous characterization has been documented to generate an overall stiffer strain response than the physiologic condition and more complex setup arrangements [21]. Consequently, this model is prone to exhibit more pronounced localized strain “hotspots” than might be physiologically accurate. Digital image correlation was used as a supplemental data acquisition technique to fill the voids in data between individual strain gage measurement locations. Statistically analyzed data in this study were exclusively gathered through standard strain gage techniques, and as a result, only reflects a small subset of the actual strain response across the entirety of the analog specimen. Results from DIC analysis indicate substantial variation in strain distribution across the pelvis in most loading scenarios. This substantial variation in strain response, even in well-distributed loading from compliant device designs, is reflected in strain gage data through high variability at each strain gage location. In addition, nonimplanted initial strain measurements were not taken, as the authors felt that a lack of a physiologic articular surface at the acetabulum prohibited sufficiently accurate unimplanted loading characteristics. The lack of baseline measurements in the unimplanted hemipelvis may limit the study of actual stress shielding characteristics. Likewise, as an analog model, the longterm biological remodeling response of the living pelvis cannot be replicated. Nevertheless, the tightly controlled nature of this setup, using a repeatable composite hemipelvis setup, allows for the direct comparison of implant alignment, JRF orientation, and acetabular component design on the change of pelvic strain response in a consistent specimen model. The primary focus of this study was to establish a simplified pelvic loading model allowing quantification and anatomical mapping of periacetabular strains, while simultaneously enabling the quantification of the change in cortical loading as a result of varied surgical and prosthetic parameters. The experimental model in this study used composite hemipelvis models from a family of analog test specimens validated against intact cadaveric bone tissues [38,39] and responded in a repeatable manner when tested for sensitivity through varying JRF vectors. Although this model does not offer the robust characteristics of an advanced computational
model nor the physiologic accuracy of a whole-pelvis cadaveric model, it does provide a uniform and consistent test medium for the preliminary comparison of relative changes in strain response due to alterations in prosthetic design and alignment. In summary, this study generated a simplified mechanical model, sensitive to loading, orientation, and component stiffness, to effectively compare the relationship between surgical and prosthetic factors and their role in cortical strain distribution in the total hip arthroplasty pelvic model. Strain gage and DIC data analysis techniques identified global implant stiffness as the greatest contributor to strain distribution across the periacetabular regions of the pelvis at the time of implantation, with stiffer acetabular components generating elevated strains at the acetabular rim in stiffer components and relatively more even distribution in the more compliant cup designs. Strain response due to orientation of the applied JRF and component abduction were less apparent near the acetabular rim, with most significant response changes occurring nearer to the body of the ilium as well as the medial wall. As the market for primary and revision THA continues to expand, a full understanding of the surgical and designspecific implications on pelvic loading, stress shielding, and the preservation of bone stock is imperative. Further study in the area of pelvic loading and stress shielding is needed to support the long-term success of cementless THA.
References 1. McLeish RD. Abduction forces in the one-legged stance. J Biomechanics 1970;3:191. 2. Hodge WA, Carlson KL, Fijan RS, et al. Contact pressures from an instrumented hip endoprosthesis. J Bone Joint Surg 1989;71:1378. 3. Bergmann G, Graichen F, Rohlmann A. Hip joint loading during walking and running, measured in two patients. J Biomechanics 1993;26:969. 4. Olson SA, Bay BK, Hamel A. Biomechanics of the hip joint and effects of fracture of the acetabulum. Clin Orthop Relat Res 1997;338:92. 5. Pedersen DR, Brand RA, Davy DT. Pelvic muscle and acetabular contact forces during gait. J Biomechanics 1997;30:959. 6. Charnley J. Total hip replacement by low-friction arthroplasty. Clin Orthop Relat Res 1970;72:7. 7. Muller ME. Total hip prosthesis. Clin Orthop Relat Res 1970;72:46. 8. Coventry MB, Beckenbaugh RD, Nolan DR, et al. 2,012 total hip arthroplasties. A study of postoperative course and early complications. J Bone Joint Surg Am 1974;56:273. 9. Harris WH. Advances in surgical technique for total hip replacement: without and with osteotomy of the greater trochanter. Clin Orthop Relat Res 1980;146:188. 10. Harkess JW. Arthroplasty of hip: surgical approaches and techniques. In: Canale SD, editor. Campbell's Operative Orthopaedics. 10th ed. St Louis, Mo: Mosby; 2003; p. 348.
Acetabular Cup Stiffness and Implant Orientation Small et al 11. Lewinnek GE, Lewis JL, Tarr R, et al. Dislocations after total hip replacement arthroplasties. J Bone Joint Surg Am 1978;60:217. 12. D'Lima DD, Urquhart AG, Buehler KO, et al. The effect of the orientation of the acetabular and femoral components on the range of motion of the hip at different head-neck ratios. J Bone Joint Surg Am 2000;82:315. 13. Dorr LD, Wan Z. Causes of and treatment protocol for instability of total hip replacement. Clin Orthop Relat Res 1998;355:144. 14. Paterno SA, Lachiewicz PF, Kelley SS. The influence of 490 patient-related factors and the position of the acetabular component on the Q9 rate of dislocation. J Bone Joint Surg Am 1997;78:163. 15. Sexton SA, Yeung E, Jackson MP, et al. The role of patient factors and implant position in squeaking of ceramic-onceramic total hip replacements. J Bone Joint Surg Br 2011; 93-B:439. 16. Walter WL, O'toole GC, Walter WK, et al. Squeaking in ceramic-on-ceramic hips: the importance of acetabular component orientation. J Arthroplasty 2007;22:496. 17. Anderson AE, Peters CL, Tuttle BD, et al. Subject-specific finite element model of the pelvis: development, validation and sensitivity studies. J Biomech Eng 2005;127:364. 18. Dalstra M, Huiskes R, van Erning L. Development and validation of a three-dimensional finite element model of the pelvic bone. Transactions of the ASME. J Biomech Eng 1995;117:272. 19. Dalstra M, Huiskes R. Load transfer across the pelvic bone. J Biomech 1995;28:715. 20. Oonishi H, Isha H, Hasegawa T. Mechanical analysis of the human pelvis and its application to the artificial hip joint— by means of the three dimensional finite element method. J Biomech 1983;16:427. 21. Phillips ATM, Pankaj P, Howie CR, et al. Finite element modeling of the pelvis: inclusion of muscular and ligamentous boundary conditions. Med Eng Phys 2007;29:739. 22. Thompson MS, Northmore-Ball MD, Tanner KE. Effects of acetabular resurfacing component material and fixation on the strain distribution in the pelvis. Proc Inst Mech Eng H 2002;216:237. 23. Vasu R, Carter DR, Harris WH. Stress distributions in the acetabular region—I. Before and after joint replacement. J Biomech 1982;15:155. 24. Carter DR, Vasu R, Harris WH. Stress distributions in the acetabular region—II. Effects of cement thickness and metal backing of the total hip acetabular component. J Biomech 1982;15:165.
367
25. Levenston ME, Beaupre GS, Schurman DJ, et al. Computer simulations of stress-related bone remodeling around noncemented acetabular components. J Arthroplasty 1993;8:595. 26. Huiskes R. Finite element analysis of acetabular reconstruction: noncemented threaded cups. Acta Orthop 1987; 58:620. 27. Dickinson AS, Taylor AC, Browne M. The influence of acetabular cup material on pelvis cortex surface strains, measured using digital image correlation. J Biomech 2012; 45:719. 28. Lionberger D, Walker PS, Granholm J. Effects of prosthetic acetabular replacement on strains in the pelvis. J Orthopaedic Research 1985;3:372. 29. Slemon D, O'Donnell P, Little E. The use of photoelastic coatings to investigate the pelvic strain variations resulting from the introduction of cementless metal backed acetabular cups. Strain 1991;27:143. 30. Mjoberg B. The theory of early loosening of implants. Orthopaedics 1997;20:1169. 31. Schmidt R, Muller L, Kress A, et al. A computed tomography assessment of femoral and acetabular bone changes after total hip arthroplasty. Int Orthop 2002; 26:299. 32. Wright JM, Pellicci PM, Salvati EA, et al. Bone density adjacent to press-fit acetabular components. A prospective analysis with quantitative computed tomography. J Bone Joint Surg 2001;83-A:529. 33. Zhang Q, Wang J, Lupton C, et al. A subject-specific pelvic bone model and its application to cemented acetabular replacements. J Biomechanics 2010;43:2722. 34. Manley MT, Ong KL, Kurtz SM. The potential for bone loss in acetabular structures following THA. Clin Orthop Relat Res 2006;453:246. 35. Meneghini RM, Ford KS, McCollough CH, et al. Bone remodeling around porous metal cementless acetabular components. J Arthroplasty 2010;25:741. 36. Pitto RP, Bhargava A, Pandit S, et al. Retroacetabular stress-shielding in THA. Clin Orthop Relat Res 2008; 466:353. 37. Stepniewski AS, Egawa H, Sychterz-Terefenko C, et al. Periacetabular bone density after total hip arthroplasty: a postmortem analysis. J Arthroplasty 2008;23:593. 38. Heiner AD. Structural properties of fourth-generation composite femurs and tibias. J Biomech 2008;41:3282. 39. Gardner MP, Chong ACM, Pollock AG, et al. Mechanical evaluation of large-size fourth-generation composite femur and tibia models. Ann Biomed Eng 2010;38:618.