M. E. Chaudhary^ Department of Orthopaedic Surgery, NYU Hospital for Joint Diseases, 301 East 17th Street, Suite 1500, New York, NY 10003; Department of Chemicai and Biomoiecuiar Engineering, NYU Polytechnic School of Engineering, Brooklyn, NY 11201 e-mail: [email protected]
P. S. Walker Department of Orthopaedic Surgery, NYU Hospitai for Joint Diseases, New York, NY Department of Mechanicai and Aerospace Engineering, NYU Polytechnic School of Engineering, Brooklyn, NY 11201
Analysis of an Early Intervention Tibial Component for Medial Osteoarthritis Tibial component loosening is an important failure mode in unicompartmental knee arthroplasty (UKA) which may be due to the 6-8 mm of bone resection required. To address component loosening and fixation, a new early intervention (El) design is proposed which reverses the traditional material scheme between femoral and tibial components. The El design consists of a plastic inlay for the distal femur and a thin metal plate for the proximal tibia. With this reversed materials scheme, the El design requires minimal tibial bone resection compared with traditional UKA. This study investigated, by means of finite element (FE) simulationst the advantages of a thin metal tibial component compared with traditional UKA tibial components, such as an all-plastic inlay or a metal-backed onlay. We hypothesized that an El tibial component would produce comparable stress, strain, and strain energy density (SED) characteristics to an intact knee and more favorable values than UKA components, due primarily to the preseiyation of dense cancellous bone near the surface. Indeed, FE results showed that stresses in the supporting bone for an El design were close to intact, while stresses, strains, and strain energy densities were reduced compared with an all-plastic UKA component. Analyzed parameters were similar for an El and a metal-backed onlay, but the El component had the advantage of minimal resection of the stiffest bone. [DOI: 10.1115/1.4027467] Keywords: osteoarthritis, tibia, unicompartmental knee, tibial stress, tibial strain
Introduction Early clinical results initially reported high failure rates for UKA with short-term follow-up [1,2], However, higher success rates have been achieved with better surgical techniques, stringent patient selection, improved surgical instrumentation, and enhanced UKA component designs. Advantages of UKA include decreased risk of infection [3], less postoperative pain [4], near restoration of normal knee kinematics [5], preservation of bone stock, and a shorter rehabilitation period compared with total knee arthroplasty. However, tibial loosening in UKA continues to be an important failure mode, while revision is not always easy due to the amount of tibial bone resected [6,7]. To address these problems, alternatives to traditional UKA component designs can be considered. This study investigated the feasibility of an El concept for the treatment of early medial osteoarthritis (OA), which is an even more conservative treatment compared with traditional UKA. Unlike UKA components, the proposed design uses a reversed materials scheme consisting of a plastic inlay component for the distal femur and a thin all-metal plate for the proximal tibia requiring only a small amount of bone resection. To investigate this reversed material scheme, we focused on the El tibial component in this study and used finite element simulations to analyze the stress, strain, and SED in the supporting cancellous bone. Results for the El tibial design were compared with existing UKA tibial designs, such as an all-plastic inlay and metal-backed onlay. All components were compared with data from the intact tibia. The parameters analyzed in underlying bone were used to assess whether an El design can provide stronger component fixation compared with UKA designs. We hypothesized that the proposed all-metal tibial design would produce comparable stress, strain.
Corresponding author. Manuscript received August 20, 2013; final manuscript received April 5, 2014; accepted manuscript posted April 22, 2014; published online May 7, 2014. Assoc. Editor: Paul RuUkoetter.
Journal of Biomechanical Engineering
and SED characteristics to the intact tibia and more favorable values than UKA components which involve greater levels of bone resection.
Methods and Materials Geometry Analyzed. From a series of UKA cases, a representative preoperative CT scan of a knee joint was obtained from the left knee of a 60 year-old male patient who was later treated with UKA. The CT scan (Siemens, 120 kVp, 1mm increment) was imported into MIMICS 16.0 software (Materialise, Leuven, Belgium), segmented slice-by-slice, and rendered to create a threedimensional model of the proximal tibia, MIMICS software was used to surgically place each tibial component in the bone model. Three design variations of the El metal tibial component were analyzed in this study (Fig. \{a)). The first had a component thickness of 2 mm at the lowest point of the dished surface; the second was the same 2 mm component with an anterior-posterior keel on the base, while the third component had a thickness of 3 mm at the lowest point. The bone resections for the El components were bi-angulated in both the coronal and sagittal planes to match the slope of the bone and to minimize the amount of dense bone resected. Components were placed at a 2 mm resection depth below the bone surface. The 2 mm keeled design was included to demonstrate that the El 2 mm design could be stiffened without increasing its thickness and hence the keeled design was compared with the El 3 mm component in this study. The El tibial component has the potential to be used as a cementless component or alternatively, fixed using a cement layer. In this study. El components were modeled in two different modes: a cementless and a cemented mode with a 1 mm thick cement mantle. In the cemented models, the 2 mm resection depth was maintained and the components were elevated 1 mm to account for the cement layer. The fourth component analyzed was an all-plastic UKA inlay of 6 mm thickness. The inlay was positioned at a 4 mm resection
Copyright © 2014 by ASME
JUNE2014, Vol. 136 / 061008-1
depth and sized to preserve 2-3 mm of peripheral bone. The allplastic inlay was angulated 3 deg from the horizontal to replicate normal alignment of the medial side. A 1 mm cement layer was included to simulate a standard implanting procedure for inlays. The last component was a UKA metal-backed onlay consisting of a 6 mm plastic bearing and a 2 mm metal base bonded together. The metal-backed onlay was placed at a 6 mm resection depth and positioned perpendicular to the long axis of the tibia to correct the varus deformity. Similar to the all-plastic inlay model, the metalbacked onlay model included a 1 mm cement layer. Mesh Sensitivity Analysis. A mesh sensitivity analysis was performed on the generated bone mode using ABAQUS/CAE 6.12 (Dassault Systems, Concord, MA). Mesh sizes of 4.5mm-1.5mm were analyzed. The model was first meshed with 4.5 mm tetrahedral elements (10-node). The elements were assigned material properties based on equations by Rho et al. [8] who correlated CT data of cancellous bone in the proximal tibia to density and elastic modulus. Equations were validated in a separate experimental study by Gray et al. [9]. Mean von Mises stresses were assessed at 2-6 mm beneath the bone surface on the medial side. Convergence was reached at a 2 mm mesh size (390,000 elements in generated models). Contact and Loading. For all cemented models, the cement layer was perfectly bonded to the tibial components and underlying bone surface. Cementless El components were modeled with bonded contact and unbonded contact between bone-implant interfaces. Bonded contact was chosen for the cementless models to simulate complete osseointegration. An initial unbonded situation was simulated using surface to surface contact and a coefficient of friction of 0.98. This friction value was taken from specification data for a rough, highly porous surface suitable for osseointegration [10]. The metal (cobalt-chrome alloy) was modeled as a linear-elastic, isotropic material (p = 8,200kg/m^ [11], £ = 208,OOOMPa [12], ^ = 0.3 [12]). Likewise the plastic, UHMWPE, was modeled as a linear-elastic, isotropic material (p = 940kg/m^ [13], £ = 634 MPa [14], ;^ = 0.45 [15]). PMMA bone cement was assigned a modulus of 2000 MPa [16] and Poisson's ratio of 0.4 [17]. Compressive loads applied to the medial and lateral condyle were taken from data of instrumented knee replacements [18]. Implanted and intact models were loaded in the central, anterior, and posterior regions, in vivo kinematic studies of unicompartmental knee replacements conducted by Argenson et al. and Banks et al. [19] and [20] have shown that net anterior-posterior femoral translation is minimal (< 5 mm). In this study, axial loading was shifted anteriorly and posteriorly ±7 mm from the center loading condition to analyze possible contact locations in extension and full flexion. On the medial side, a downward force of 1500 N was applied in the direction of the long axis of the tibia over an area of 40 mm" which is a representative contact area for metal on plastic components during walking [21]. The medial load was applied directly to the surface of the tibial component. On the intact lateral side, a downward force of 750 N was applied over an area of 450 mm^ based on data using knee specimens [22]. The lateral load was applied to the surface of the tibia. An upward force of 1500N was applied over the tibial tubercle to simulate the force exerted by the patella tendon. In all models, the base of the tibia was fully constrained. Only central loading was applied to an El 2 mm component with keel. The keel design was included as a preliminary analysis for comparison to keel-less El designs. Reference Intact Model. The intact tibia bone was used as a base reference to compare with the implanted models. The same axial forces were applied, however, the contact area on the 061008-2 / Vol.136, JUNE 2014
articular suriace of the medial side was 554 mm^ to represent loading in an intact knee. This value was measured from the MRI scans of five cadaveric knee specimens subjected to an axial compression force in a separate study [23]. Loads for the intact model were applied directly to the tibial plateau. The intact tibial model was cut at resection depths of 2 mm, 4 mm, and 6 mm resulting in three different cases, with each case composed of two regions: a cut plateau region (2, 4, 6 mm thick) and a main tibial bone region. For simulation purposes, these two regions were bonded together. However, the resected regions were removed during postprocessing to compare stress, strain, and SED parameters at various resection depths against the implanted models. Parameters Analyzed. Von Mises stresses, principal strains (maximum and minimum), and SED were obtained near the resected surface for implanted and intact models. Regarding the significance of the stress and strain parameters calculated, stresses at the bone interface can be used to assess load distribution [24], while the interface strains indicate possible loosening and bone failure. SED below the components can be used as an indication of bone remodeling and/or potential pain [25,26]. The region of interest for this study was the resected bone surface and trabecular bone a few millimeters just below the resected surface based on the assumption that high stresses, strains, and SED at the interface and a few millimeters below might lead to bone failure, migration and hence, loosening. Within the region of interest, mean maximum values were calculated for each parameter by sorting values within the region from largest to smallest and averaging the largest ten values. For bone tissue, yield strains in tension and in compression were estimated to be 0.65% and 0.73%, respectively [27].
Results The change in bone density and modulus with respect to resection depth in the proximal tibia is illustrated in Fig. \(b). At a 2 mm resection depth on the medial side, there is a concentration of high density and high modulus bone near the center in the anterior-posterior direction. An El component would be positioned at a 2 mm resection. UKA components are typically placed at 4 mm-8 mm of resection. As resection depths increase to 8 mm in increments of 2 mm, the density of the trabecular bone rapidly decreases. At a 4 mm resection, the maximum modulus in the central trabecular region of the medial condyle decreased 20.8% when compared with the maximum modulus in the same area at a 2 mm resection. Similarly, the maximum modulus at a 6 mm resection decreased 24.4% when compared with the maximum modulus at a 4 mm resection. Lastly, the maximum modulus at an 8 mm resection decreased 29.7% when compared with the maximum modulus at a 6 mm resection. Figure 2 compares mean maximum von Mises stresses between El components, UKA components, and an intact model at central, anterior, and posterior loading positions. Components were analyzed at different resection levels and stress values for an intact tibia are shown at various resection depths. In Fig. 2(a), stresses
Deniitv
I
""7äöi"
I
240.21 410.52 S80.&2 750.82 92102 1091.22 Ï261.42 "14Î.62 1M1.S3
E. Modulus 171.97 931.09 193949 3119.e 443Ë.6) S369.74 74tM.7S 9031.28 10741.27 Í2SZS.3J
Fig. 1 (a) Five components analyzed: El 2 mm, El 2 mm with keel. El 3 mm, all-plastic UKA, metal-backed UKA. (b) Change in bone density (kg/m^) and elastic modulus (MPa) with resection depths ranging from 2 mm to 8 mm. Transactions of the ASME
I 7 Diitsnce Uom center (mm) Neg«ttve • Pof terlor Potltlve Anterior —»-tnlactat 2 m —•—IniMtatfim .i#ii t l 2nimcerr«ntlea (urdwnded) —•—£l3mmc«ni «ntleu (bonded) —•—£13mm w/c mwnt - • - I n b v w / c « ! » em
ntleu (bonded) -•^~EI2mm w/ce rn«nt —•—CrSmmtem n l l ^ u (unbonded) mt
Fig. 2 (a) Maximum von Mises stresses for center and offset loading in the anterior-posterior direction, (b) Contour plots of von Mises stresses at the bone interface for cemented components centrally loaded. for an El 3 mm and the metal-backed onlay were similar to an intact tibial model centrally loaded. The onlay showed larger differences in magnitude and distribution when loaded anteriorly and posteriorly. At center loading, stresses for both cemented El 2 mm and all-plastic inlay components (10.14 MPa and 19.4 MPa, respectively) were higher than an intact model (8.6 MPa at a 2 mm depth and 8.4 MPa at a 4 mm depth). Bone stresses were contour plotted for al] components (Fig. 2(b)). High stress values were found beneath the load, corresponding loosely to regions of stiff bone. Cementless El 2 mm components (bonded and unbonded) produced stresses slightly higher than an intact model when loaded centrally. Stresses were closer to intact when loaded anteriorly and posteriorly. Figure 3 compares maximum and minimum principal strains between tibial components analyzed in this study and an intact tibia model. When loaded centrally, anteriorly, or posteriorly, strain values corresponding to an El 2 mm cemented component. El 3 mm cemented component, and metal-backed onlay were slightly lower than strain values calculated for an intact model (Fig. 3(a)). Strains for an all-plastic inlay were much higher compared with an intact model and the El components for all loading conditions. Strains for all components were elevated when loaded posteriorly. When loaded anteriorly, strains for all components
Fig. 4 (a) Maximum strain energy densities for center and offset loading in the anterior-posterior direction, (b) Contour plots of strain energy densities within the bone for cemented components centrally loaded. except the all-plastic inlay decreased slightly when compared with central loading. Strains for cementless El components were close to intact strain values. Unbonded cementless El components generally produced slightly higher strain values than bonded cementless components. Strain distributions for central loading were contour plotted in Fig. 3(b). High strains near the lateral edge of the resected bone were caused in part by the comer of the cement mantle or component edge and low bone stiffness in that region (Fig. l(a)). For this analysis, strain values were analyzed at the relative center in the mediolateral direction. The strain values calculated for all models did not exceed the estimated yield strain of natural bone (0.65% in tension, 0.73% in compression) [27]. Compared with an intact model, the SED for a cemented El 3 mm component as well as a metal-backed UKA decreased by 57% and 36%, respectively. Conversely, the SED for an allplastic inlay UKA compared with an intact tibia increased by 332%. The SED distribution of an all-plastic inlay, positioned at a 4 mm resection depth, shows a high concentration of SED values directly below the component. Whereas SED distributions for an El 2 mm. El 3 mm, and a metal-backed onlay were very similar to the SED of an intact model (Fig. 4(è)). SED for cementless El 2 mm components (bonded and unbonded) was slightly higher than intact SED values. Stresses and strains for an El 2 mm with a keel design slightly decreased compared with an El 3 mm component with a flat backing at the same resection level (Fig. 5). However, much higher stresses and strains were observed in the bone region at the base of the keel. In an El 2 mm with keel model at the bone interface, stresses and minimum principal strains decreased by 14% and 15%, respectively, compared with an El 3 mm model while maximum principal strains increased 76%.
El 2 mm with keel
Intact al 2 mm
Ohtance from center (mm) Nesatlve > Posterior PoiHIve • Anlerfor — • - i m s c t at
len (unbondnl) El ïmmcementls El 3mm w/ Metal tMdced onlay
El 2 mm with keel
Fig. 3 (a) Maximum and minimum principai strains for center and offset ioading in the anterior-posterior direction, (b) Contour plots of maximum and minimum principai strains at the bone interface for cemented components centraily ioaded. Journal of Biomechanical Engineering
Fig. 5 Contour piots of stresses and strains at the bone interface surface of an Ei 3 mm component and El 2 mm component with keel design JUNE2014, Vol. 136 / 061008-3
Discussion In this study, we presented a novel approach for the treatment of early OA called the early intervention design which reverses the material scheme between femoral and tibial components compared with traditional UKA components. Such a scheme was introduced by Charnley [28] in the mid 1970s but was limited in use. Using today's materials and techniques, the concept may well be feasible. With a reversed materials scheme, the El design requires minimal tibial bone resection which preserves the dense and stiff bone near the surface of the proximal tibia. FEA methods were used to investigate the stress, strain, and SED differences between El tibial components, an intact tibia model, and traditional UKA tibial components. The all-plastic inlay showed elevated stresses, strains, and SED compared with an intact reference model. El 2 mm. El 3 mm, and a metal-backed onlay. For the inlay, the elevated SED, stress, and strain may be indicators of loosening, bone remodeling, and pain reported clinically [26,29-31] caused in part by the low modulus of the plastic and its thickness. Although FE simulations used idealized assumptions, such as even cement layers and complete bonded contact between surfaces, an all-plastic inlay component produced very high strains which may explain its tendency to fail early as shown in published experimental and clinical studies. Experimental studies have reported higher strains associated with inlay designs compared with metal-backed onlays using implanted tibia models [31,32], while clinically, inlays have been shown to fail much sooner than metal-backed onlay designs [33,34]. The causes of tibial component subsidence and loosening are still largely unknown; however, it is believed that a number of factors, such as bone strength and overall design of a tibial component, can lead to subsidence. These two factors also affect strains within the supporting bone as demonstrated in this study. For an El 2 mm, the small thickness most likely caused the elevated stresses, strains, and SED. A cement layer helped to reduce stresses and strains in the supporting bone [35]. Similarly, with the addition of a central keel, the El 2 mm showed reduced stress and strain values when compared with an El 3 mm and metalbacked onlay. The keel distributed the load more effectively onto the underlying bone by stiffening the center of the component where the load is applied. Additionally, dishing the upper surface of the El design further stiffened the component. In terms of preserving strong bone, low resection depths are an advantage. The bone strength and density have been found to diminish with depth below the surface, in both normal [36] and osteoarthritic cases [37,38]. An El 3 mm and metal-backed onlay produced comparable stresses and strains when loaded centrally, anteriorly, and posteriorly; however, an El 3 mm component only requires a 2 mm resection, whereas a metal-backed onlay requires a 6-8 mm resection. With less resection, the stronger and denser bone would be preserved providing stronger fixation of the component and ease of a future revision surgery since more bone stock is available. In this study, we used a typical arthritic tibia with a nonuniform distribution of bone properties in the proximal tibia. The importance of using such a bone model was that it affected the stress and strain distributions and took account of the levels of resection required for each type of component. Additional evaluation parameters will now be necessary to test the viability of the concept in the tibia. From a practical point of view, in cemented components, cement penetration will be restricted in dense bone, and hence experiments will be necessary using techniques, such as microdrilling. We analyzed one tibial sample, but in OA, the bone density patterns vary widely [37,38], requiring analyses which cover the range. Axial loads were applied centrally, anteriorly, and posteriorly to all models analyzed here, however, other force components representing a range of loading scenarios need to be further investigated. This is particularly important if El components are to be 061008-4 / Vol. 136, JUNE 2014
used uncemented where interface stability is essential for osseointegration, although it should be noted that for offset loading in the bonded condition, there were minimal tensile stresses found across the interface. For the cementless El components, minimal differences in stress, strain, or SED were found between bonded and unbonded cases. For cementless El components, it will be desirable to have a very rough underside for initial stability [10]. The femoral component also needs analysis, using, for example, the configuration of Charnley, a 6-8 mm plastic inlay embedded in the distal femur, consistent with the osteoarthritic lesion in early OA [39]. This component could be all-pla.stic or have a metal backing. In either case, even if it had to be revised to a total knee in the future, this would involve 10 mm of distal femoral resection, completely containing the implant. In summary, FE simulations detnonstrated the effectiveness of using an early intervention tibial component over traditional UKA tibial components. The El design will require small incisions and bone resections, which will allow for natural anatomy of the patient to be preserved and less invasive surgery techniques to be used. This study showed that an El design has the potential to reduce the loosening rates seen in contemporary UKA designs.
Acknowledgment The Department of Orthopaedic Surgery, New York University, Hospital for Joint Diseases funded this study. We would like to thank Professor Nikhil Gupta and Dr. Oran Kennedy for helpful discussions. We are also grateful to Dr. Joseph Bosco and Dr. Ivan Fernandez-Madrid from New York University, Hospital for Joint Diseases for their valuable medical perspective regarding component feasibility. This study was selected for the ISTA 2013 Student Biomechanics Award.
References [1] Insall, J., and Aglietti, P., 1980, "A Five to Seven-Year Follow-Up of Unicondylar Arthroplasty," J. Bone Joint Surg. Am., 62, pp. 1329-1337. [2] Swank, M., Stulberg, S. D., Jiganti, J.. and Machairas, S.. 1993. "The Natural History of Unicompartmental Arthroplasty. An Eight-Year Follow-Up Study With Survivorship Analysis," Clin. Orthop. Relat. Res., 286, pp. 130-142. [3] Cheng, T., Guoyou, Z., and Zhang, X., 2010, "Does Minimally Invasive Surgery Improve Short-Term Recovery in Total Knee Arthroplasty?," Clin. Orthop. Relat. Res., 468, pp. 1635-1648. |4] McAllister, C , and Stepanian, J., 2008, "The Impact of Minimally Invasive Surgical Techniques on Early Range of Motion after Primary Total Knee Arthroplasty," J. Arthroplasty, 23(1), pp. 10-18. 15] Khanna, A., Gougoulias, N., Longo U. G., and Maffulli, N., 2009, "Minimally Invasive Total Knee Arthroplasty: A Systematic Review." Orthop. Clin. North Am., 40(4), pp. 479^89. |6] Borus, T., and ThornhiU, T., 2008, "Unicompartmental Knee Arthroplasty," J. Am. Acad. Orthop. Surg., 16(1). pp. 9-18. [71 Tanavalee, A., Choi, Y. J., and Trai, A. J., 2005, "Unicondylar Knee Arthroplasty: Past and Present," Orthopedics, 28, pp. 1423-1433. [8) Rho, J. Y., Hobatho, M. C , and Ashman, R. B., 1995. "Relations of Mechanical Properties to Density and CT Numbers in Human Bone," Med. Eng. Phys., 17(5), pp. 347-355. [91 Gray, H., Taddei, F., Zavatsky, A., Cristofolini, L., and Gill, H., 2008, "Experimental Validation of a Finite Element Model of a Human Cadaveric Tibia," ASME J. Biomech. Eng., 130, p. 031016. [10] Zhang, Y., Ahn, P. B., Fitzpatrick. D. C, Heiner, A., Poggie, R. A., and Brown, T. D., 1999, "Interfacial Frictional Behavior; Cancellous Bone, Cortical Bone, and a Novel Porous Tantalum Biomaterial," J. Musculoskeletal Res., 3(4), pp. 245—251. [11] Tandon. R., 1999, "Net Shaping of Co-Cr-Mo (F-75) via Metal Injection Molding," Cohall-Base Alloys for Biomédical Applications, J. A. Disegi, R. L. Kennedy, and R. Pilliar, eds., ASTM International, West Conshohocken, PA, pp. 3-10. [12) Rapperport, D. J., Carter, D. R., and Schurman, D. J., 1987, "Contact Finite Element Stress Analysis of Porous Ingrowth Aeetabular Cup Implantation, Ingrowth, and Loosening," J. Orthop. Res., 5(4), pp. 548-561. [13] Shanbhag, A. S., Hasselman, C. T., and Rubash, H. E., 1996, "Technique for Generating Submicrometer Ultra High Molecular Weight Polyethylene Particles," J. Orthop. Res., 14, pp. 1000-1004. [14] Kurtz, S. M.. Villarraga, M. L.. Herr, M. P., Bergström, J. S., and Rimnac, C. M., 2002, "Thermomechanical Behavior of Virgin and Highly Crosslinked Ultra-High Molecular Weight Polyethylene Used in Total Joint Replacements," Biomaterials, 23(17), pp. 3681-3697. [15] Ho, S. P., Carpick. R., Boland, T., and LaBerge, M., 2002, "Nanotribology of Cocr-UHMWPF TJR Prosthesis Using Atomic Force Microscopy," Wear, 253(11-12), pp. 1145-1155.
Transactions of the ASME
[16] Lakes, R. S., 2002, Composite Biomaterials. Biomaterials: Principles and Applications, CRC Press, LLC, Boca Raton, FL, p. 84. [17] Orr, J. F., Dunne, N. J., and Quinn, Q. C, 2003, "Shrinkage Stresses in Bone Cement," Biomaterials, 24(17), pp. 2933-2940. [18] Mündemiann, A., Dyrby, C. O., D'Lima, D. D.. Colwell, C. W., Jr., and Andriacchi, T. P., 2008, "in vivo Knee Loading Characteristics During Activities of Daily Living as Measured by an Instrumented Total Knee Replacement," J. Orthop. Res., 26(9), pp. 1167-1172. [19] Argenson, J. A., Komistek, R. D., Aubaniac, J., Dennis, D. A., Northcut, E., Anderson, D., and Agostini, S., 2002, "In Vivo Determination of Knee Kinematics for Subjects Implanted With a Unicompartmental Arthroplasty," J. Arthroplasty, 17(8), pp. 1049-1054. [20] Banks, S., Fregly, B., Boniforti, F., Reinschmidt, C, and Romagnoli, S., 2005, "Comparing In Vivo Kinematics of Unicondylar and Bi-Unicondylar Knee Replacements," Knee, 13, pp. 55Í-556. [21] D'Lima, D. D., Patil, S., Steklov, N., and Colwell, C. W., 2011, "The 2011 ABJS Nicolas Andry Award: 'Lab'-In-A-Knee: in vivo Knee Forces, Kinematics, and Contact Analysis," Clin. Orthop. Relat. Res., 469(10), pp. 2953-2970. [22] Fukubayashi, T.. and Kurosawa, H., 1980, "The Contact Area and Distribution Pattem of the Knee," Acta Orthop. Scand., 51, pp. 871-879. [23] Bell, C , Amo, S., Hadley, S., Campbell, K., Beltran L., Recht M., Sherman O., Walker P., 2012. "Relative Contacts Between the Medial Meniscus and Exposed Cartilage in the Knee," 58th Annual Orthopaedic Research Society. [24] Taylor, M., Tanner, K. E., and Freeman, M. A. R., 1998, "Finite Element Analysis of the Implanted Proximal Tibia: A Relationship Between the Initial Cancellous Bone Stresses and Implant Migration," J. Biomech., 31, pp. 303-310. [25] Huiskes, R., Weinans, H., Grootenboer, H. J., Dalstra, M., Fudala, B., and Sloof, T. J., 1987, "Adaptive Bone-Remodeling Theory Applied to ProstheticDesign Analysis," J. Biomech., 20(11-12), pp. 1135-1150. [26] Simpson. D., Kendrick, B., Dodd, C, Price, A., Gill, H., and Murray, D., 2010, "Load Transfer in the Proximal Tibia Following Implantation With a Unicompartmental Knee Replacement: A Static Snapshot," Proc. Inst. Mech. Eng., Part H, 225, pp. 521-529. [27] Morgan, E. F., and Keaveny, T. M., 2001, "Dependence of Yield Strain on Human Trabecular Bone on Anatomic Site," J. Biomech., 34, pp. 569—577.
Journal of Biomechanical Engineering
[28] Minas, R. J., Day, J. B., and Hardinge, K., 1982, "Kinesiologic and Biomechanical Assessment of the Chamley 'Load Angle Inlay' Knee Prosthesis," Eng. Med., 11(1), pp. 25-32. [29] Arastu, M., Vijayaraghavan, J., Chissell, H., Hull, J., Newman, J., and Robison, J., 2009, "Early Failure of a Mobile Bearing Unicompartmental Knee Replacement," Knee Surg. Sports Traumatol. Arthrose, 17(10), pp. 1178-1183. [30] Lisowski L. A., van den Bekerom M. P. J., Pilot P., van Dijk C. N., Lisowski A. E., 2011, "Oxford Phase 3 Unicompartmental Knee Arthroplasty: MediumTerm Results of a Minimally Invasive Surgical Procedure," Knee Surg Sports Traumatol Arthrose, 19(2): 277-84. [31] Small, S. R., Berend, M. E., Ritter, M. D., Buckley, C. A., and Rogge, R. D., 2011, "Metal Backing Significantly Decreases Tibial Strains in a Medial Unicompartmental Knee Arthropiasty Model," J. Arthroplasty, 26(5), pp. 777-782. [32] Scott, C. E. H., Eaton, M. J., Nutton, R. W.. Wade, F., Pankaj, P., and Evans, S., 2013, "Proximal Tibial Strain in Medial Unicompartmental Knee Replacements," Bone Joint J., 95B, pp. 1339-1347. [33] Saenz, C. L., McGrath, M. S., Marker, D. R., Seyler, T. M., Mont, M. A., and Bonutti, P. M., 2010, "Early Failure of a Unicompartmental Knee Arthroplasty Design With an All-Polyethylene Tibial Component," Knee, 17, pp. 53-56. [34] Gladnick, B., Nam, D., Khamaisy, S., Paul, S., and Pearle. A., "Inlay Versus Onlay Tibial Implants in Robotic Unicondylar Knee Arthroplasty," Annual Congress of the Intemational Society for Technology in Arthroplasty, Palm Beach, FUOct. 16-19. [35] Thompson, M. T., Conditt, M. A., Otto, J. K., and Redish, M., 2010, "The knportance of Good Cement Mantle With an All-Poly Inlay UKA" Transactions of the 56th Annual Meeting—Orthopaedic Research Society, Orthopaedic Re.search Society. [36] Hvid, 1., and Hansen, S. L., 1985, "Trabecular Bone Strength Pattems at the Proxitnal Tibial Epiphysis," J. Orthop. Res., 3(4), pp. 464-472. [37] Hvid, I., Í988, "Trabecuiar Bone Strength at the Knee," Clin. Orthop. Relat. Res., 227, pp. 210-221. [38] Wong, N., Wei, C. S.. Gautam, P., and Walker, P. S., 2012, "Multi-Planar Visualization of Tibial Bone Density Distribution From a Tibial CAT Scan Dataset," Orthopedics Research Society Annual Meeting, February, 2012. [39] Arno, S., Maffei, D., Walker, P. S., Schwarzkopf, R., Desai, P., and Steiner, G., 2011, "Retrospective Analysis of Total Knee Arthroplasty Cases for Visual, Histological and Clinical Eligibility of Unicompartmental Knee Arthroplasties," J. Arthroplasty, 26(8), pp. 1396-1403.
JUNE2014, Vol. 136 / 061008-5
Copyright of Journal of Biomechanical Engineering is the property of American Society of Mechanical Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
P. S. Walker Department of Orthopaedic Surgery, NYU Hospitai for Joint Diseases, New York, NY Department of Mechanicai and Aerospace Engineering, NYU Polytechnic School of Engineering, Brooklyn, NY 11201
Analysis of an Early Intervention Tibial Component for Medial Osteoarthritis Tibial component loosening is an important failure mode in unicompartmental knee arthroplasty (UKA) which may be due to the 6-8 mm of bone resection required. To address component loosening and fixation, a new early intervention (El) design is proposed which reverses the traditional material scheme between femoral and tibial components. The El design consists of a plastic inlay for the distal femur and a thin metal plate for the proximal tibia. With this reversed materials scheme, the El design requires minimal tibial bone resection compared with traditional UKA. This study investigated, by means of finite element (FE) simulationst the advantages of a thin metal tibial component compared with traditional UKA tibial components, such as an all-plastic inlay or a metal-backed onlay. We hypothesized that an El tibial component would produce comparable stress, strain, and strain energy density (SED) characteristics to an intact knee and more favorable values than UKA components, due primarily to the preseiyation of dense cancellous bone near the surface. Indeed, FE results showed that stresses in the supporting bone for an El design were close to intact, while stresses, strains, and strain energy densities were reduced compared with an all-plastic UKA component. Analyzed parameters were similar for an El and a metal-backed onlay, but the El component had the advantage of minimal resection of the stiffest bone. [DOI: 10.1115/1.4027467] Keywords: osteoarthritis, tibia, unicompartmental knee, tibial stress, tibial strain
Introduction Early clinical results initially reported high failure rates for UKA with short-term follow-up [1,2], However, higher success rates have been achieved with better surgical techniques, stringent patient selection, improved surgical instrumentation, and enhanced UKA component designs. Advantages of UKA include decreased risk of infection [3], less postoperative pain [4], near restoration of normal knee kinematics [5], preservation of bone stock, and a shorter rehabilitation period compared with total knee arthroplasty. However, tibial loosening in UKA continues to be an important failure mode, while revision is not always easy due to the amount of tibial bone resected [6,7]. To address these problems, alternatives to traditional UKA component designs can be considered. This study investigated the feasibility of an El concept for the treatment of early medial osteoarthritis (OA), which is an even more conservative treatment compared with traditional UKA. Unlike UKA components, the proposed design uses a reversed materials scheme consisting of a plastic inlay component for the distal femur and a thin all-metal plate for the proximal tibia requiring only a small amount of bone resection. To investigate this reversed material scheme, we focused on the El tibial component in this study and used finite element simulations to analyze the stress, strain, and SED in the supporting cancellous bone. Results for the El tibial design were compared with existing UKA tibial designs, such as an all-plastic inlay and metal-backed onlay. All components were compared with data from the intact tibia. The parameters analyzed in underlying bone were used to assess whether an El design can provide stronger component fixation compared with UKA designs. We hypothesized that the proposed all-metal tibial design would produce comparable stress, strain.
Corresponding author. Manuscript received August 20, 2013; final manuscript received April 5, 2014; accepted manuscript posted April 22, 2014; published online May 7, 2014. Assoc. Editor: Paul RuUkoetter.
Journal of Biomechanical Engineering
and SED characteristics to the intact tibia and more favorable values than UKA components which involve greater levels of bone resection.
Methods and Materials Geometry Analyzed. From a series of UKA cases, a representative preoperative CT scan of a knee joint was obtained from the left knee of a 60 year-old male patient who was later treated with UKA. The CT scan (Siemens, 120 kVp, 1mm increment) was imported into MIMICS 16.0 software (Materialise, Leuven, Belgium), segmented slice-by-slice, and rendered to create a threedimensional model of the proximal tibia, MIMICS software was used to surgically place each tibial component in the bone model. Three design variations of the El metal tibial component were analyzed in this study (Fig. \{a)). The first had a component thickness of 2 mm at the lowest point of the dished surface; the second was the same 2 mm component with an anterior-posterior keel on the base, while the third component had a thickness of 3 mm at the lowest point. The bone resections for the El components were bi-angulated in both the coronal and sagittal planes to match the slope of the bone and to minimize the amount of dense bone resected. Components were placed at a 2 mm resection depth below the bone surface. The 2 mm keeled design was included to demonstrate that the El 2 mm design could be stiffened without increasing its thickness and hence the keeled design was compared with the El 3 mm component in this study. The El tibial component has the potential to be used as a cementless component or alternatively, fixed using a cement layer. In this study. El components were modeled in two different modes: a cementless and a cemented mode with a 1 mm thick cement mantle. In the cemented models, the 2 mm resection depth was maintained and the components were elevated 1 mm to account for the cement layer. The fourth component analyzed was an all-plastic UKA inlay of 6 mm thickness. The inlay was positioned at a 4 mm resection
Copyright © 2014 by ASME
JUNE2014, Vol. 136 / 061008-1
depth and sized to preserve 2-3 mm of peripheral bone. The allplastic inlay was angulated 3 deg from the horizontal to replicate normal alignment of the medial side. A 1 mm cement layer was included to simulate a standard implanting procedure for inlays. The last component was a UKA metal-backed onlay consisting of a 6 mm plastic bearing and a 2 mm metal base bonded together. The metal-backed onlay was placed at a 6 mm resection depth and positioned perpendicular to the long axis of the tibia to correct the varus deformity. Similar to the all-plastic inlay model, the metalbacked onlay model included a 1 mm cement layer. Mesh Sensitivity Analysis. A mesh sensitivity analysis was performed on the generated bone mode using ABAQUS/CAE 6.12 (Dassault Systems, Concord, MA). Mesh sizes of 4.5mm-1.5mm were analyzed. The model was first meshed with 4.5 mm tetrahedral elements (10-node). The elements were assigned material properties based on equations by Rho et al. [8] who correlated CT data of cancellous bone in the proximal tibia to density and elastic modulus. Equations were validated in a separate experimental study by Gray et al. [9]. Mean von Mises stresses were assessed at 2-6 mm beneath the bone surface on the medial side. Convergence was reached at a 2 mm mesh size (390,000 elements in generated models). Contact and Loading. For all cemented models, the cement layer was perfectly bonded to the tibial components and underlying bone surface. Cementless El components were modeled with bonded contact and unbonded contact between bone-implant interfaces. Bonded contact was chosen for the cementless models to simulate complete osseointegration. An initial unbonded situation was simulated using surface to surface contact and a coefficient of friction of 0.98. This friction value was taken from specification data for a rough, highly porous surface suitable for osseointegration [10]. The metal (cobalt-chrome alloy) was modeled as a linear-elastic, isotropic material (p = 8,200kg/m^ [11], £ = 208,OOOMPa [12], ^ = 0.3 [12]). Likewise the plastic, UHMWPE, was modeled as a linear-elastic, isotropic material (p = 940kg/m^ [13], £ = 634 MPa [14], ;^ = 0.45 [15]). PMMA bone cement was assigned a modulus of 2000 MPa [16] and Poisson's ratio of 0.4 [17]. Compressive loads applied to the medial and lateral condyle were taken from data of instrumented knee replacements [18]. Implanted and intact models were loaded in the central, anterior, and posterior regions, in vivo kinematic studies of unicompartmental knee replacements conducted by Argenson et al. and Banks et al. [19] and [20] have shown that net anterior-posterior femoral translation is minimal (< 5 mm). In this study, axial loading was shifted anteriorly and posteriorly ±7 mm from the center loading condition to analyze possible contact locations in extension and full flexion. On the medial side, a downward force of 1500 N was applied in the direction of the long axis of the tibia over an area of 40 mm" which is a representative contact area for metal on plastic components during walking [21]. The medial load was applied directly to the surface of the tibial component. On the intact lateral side, a downward force of 750 N was applied over an area of 450 mm^ based on data using knee specimens [22]. The lateral load was applied to the surface of the tibia. An upward force of 1500N was applied over the tibial tubercle to simulate the force exerted by the patella tendon. In all models, the base of the tibia was fully constrained. Only central loading was applied to an El 2 mm component with keel. The keel design was included as a preliminary analysis for comparison to keel-less El designs. Reference Intact Model. The intact tibia bone was used as a base reference to compare with the implanted models. The same axial forces were applied, however, the contact area on the 061008-2 / Vol.136, JUNE 2014
articular suriace of the medial side was 554 mm^ to represent loading in an intact knee. This value was measured from the MRI scans of five cadaveric knee specimens subjected to an axial compression force in a separate study [23]. Loads for the intact model were applied directly to the tibial plateau. The intact tibial model was cut at resection depths of 2 mm, 4 mm, and 6 mm resulting in three different cases, with each case composed of two regions: a cut plateau region (2, 4, 6 mm thick) and a main tibial bone region. For simulation purposes, these two regions were bonded together. However, the resected regions were removed during postprocessing to compare stress, strain, and SED parameters at various resection depths against the implanted models. Parameters Analyzed. Von Mises stresses, principal strains (maximum and minimum), and SED were obtained near the resected surface for implanted and intact models. Regarding the significance of the stress and strain parameters calculated, stresses at the bone interface can be used to assess load distribution [24], while the interface strains indicate possible loosening and bone failure. SED below the components can be used as an indication of bone remodeling and/or potential pain [25,26]. The region of interest for this study was the resected bone surface and trabecular bone a few millimeters just below the resected surface based on the assumption that high stresses, strains, and SED at the interface and a few millimeters below might lead to bone failure, migration and hence, loosening. Within the region of interest, mean maximum values were calculated for each parameter by sorting values within the region from largest to smallest and averaging the largest ten values. For bone tissue, yield strains in tension and in compression were estimated to be 0.65% and 0.73%, respectively [27].
Results The change in bone density and modulus with respect to resection depth in the proximal tibia is illustrated in Fig. \(b). At a 2 mm resection depth on the medial side, there is a concentration of high density and high modulus bone near the center in the anterior-posterior direction. An El component would be positioned at a 2 mm resection. UKA components are typically placed at 4 mm-8 mm of resection. As resection depths increase to 8 mm in increments of 2 mm, the density of the trabecular bone rapidly decreases. At a 4 mm resection, the maximum modulus in the central trabecular region of the medial condyle decreased 20.8% when compared with the maximum modulus in the same area at a 2 mm resection. Similarly, the maximum modulus at a 6 mm resection decreased 24.4% when compared with the maximum modulus at a 4 mm resection. Lastly, the maximum modulus at an 8 mm resection decreased 29.7% when compared with the maximum modulus at a 6 mm resection. Figure 2 compares mean maximum von Mises stresses between El components, UKA components, and an intact model at central, anterior, and posterior loading positions. Components were analyzed at different resection levels and stress values for an intact tibia are shown at various resection depths. In Fig. 2(a), stresses
Deniitv
I
""7äöi"
I
240.21 410.52 S80.&2 750.82 92102 1091.22 Ï261.42 "14Î.62 1M1.S3
E. Modulus 171.97 931.09 193949 3119.e 443Ë.6) S369.74 74tM.7S 9031.28 10741.27 Í2SZS.3J
Fig. 1 (a) Five components analyzed: El 2 mm, El 2 mm with keel. El 3 mm, all-plastic UKA, metal-backed UKA. (b) Change in bone density (kg/m^) and elastic modulus (MPa) with resection depths ranging from 2 mm to 8 mm. Transactions of the ASME
I 7 Diitsnce Uom center (mm) Neg«ttve • Pof terlor Potltlve Anterior —»-tnlactat 2 m —•—IniMtatfim .i#ii t l 2nimcerr«ntlea (urdwnded) —•—£l3mmc«ni «ntleu (bonded) —•—£13mm w/c mwnt - • - I n b v w / c « ! » em
ntleu (bonded) -•^~EI2mm w/ce rn«nt —•—CrSmmtem n l l ^ u (unbonded) mt
Fig. 2 (a) Maximum von Mises stresses for center and offset loading in the anterior-posterior direction, (b) Contour plots of von Mises stresses at the bone interface for cemented components centrally loaded. for an El 3 mm and the metal-backed onlay were similar to an intact tibial model centrally loaded. The onlay showed larger differences in magnitude and distribution when loaded anteriorly and posteriorly. At center loading, stresses for both cemented El 2 mm and all-plastic inlay components (10.14 MPa and 19.4 MPa, respectively) were higher than an intact model (8.6 MPa at a 2 mm depth and 8.4 MPa at a 4 mm depth). Bone stresses were contour plotted for al] components (Fig. 2(b)). High stress values were found beneath the load, corresponding loosely to regions of stiff bone. Cementless El 2 mm components (bonded and unbonded) produced stresses slightly higher than an intact model when loaded centrally. Stresses were closer to intact when loaded anteriorly and posteriorly. Figure 3 compares maximum and minimum principal strains between tibial components analyzed in this study and an intact tibia model. When loaded centrally, anteriorly, or posteriorly, strain values corresponding to an El 2 mm cemented component. El 3 mm cemented component, and metal-backed onlay were slightly lower than strain values calculated for an intact model (Fig. 3(a)). Strains for an all-plastic inlay were much higher compared with an intact model and the El components for all loading conditions. Strains for all components were elevated when loaded posteriorly. When loaded anteriorly, strains for all components
Fig. 4 (a) Maximum strain energy densities for center and offset loading in the anterior-posterior direction, (b) Contour plots of strain energy densities within the bone for cemented components centrally loaded. except the all-plastic inlay decreased slightly when compared with central loading. Strains for cementless El components were close to intact strain values. Unbonded cementless El components generally produced slightly higher strain values than bonded cementless components. Strain distributions for central loading were contour plotted in Fig. 3(b). High strains near the lateral edge of the resected bone were caused in part by the comer of the cement mantle or component edge and low bone stiffness in that region (Fig. l(a)). For this analysis, strain values were analyzed at the relative center in the mediolateral direction. The strain values calculated for all models did not exceed the estimated yield strain of natural bone (0.65% in tension, 0.73% in compression) [27]. Compared with an intact model, the SED for a cemented El 3 mm component as well as a metal-backed UKA decreased by 57% and 36%, respectively. Conversely, the SED for an allplastic inlay UKA compared with an intact tibia increased by 332%. The SED distribution of an all-plastic inlay, positioned at a 4 mm resection depth, shows a high concentration of SED values directly below the component. Whereas SED distributions for an El 2 mm. El 3 mm, and a metal-backed onlay were very similar to the SED of an intact model (Fig. 4(è)). SED for cementless El 2 mm components (bonded and unbonded) was slightly higher than intact SED values. Stresses and strains for an El 2 mm with a keel design slightly decreased compared with an El 3 mm component with a flat backing at the same resection level (Fig. 5). However, much higher stresses and strains were observed in the bone region at the base of the keel. In an El 2 mm with keel model at the bone interface, stresses and minimum principal strains decreased by 14% and 15%, respectively, compared with an El 3 mm model while maximum principal strains increased 76%.
El 2 mm with keel
Intact al 2 mm
Ohtance from center (mm) Nesatlve > Posterior PoiHIve • Anlerfor — • - i m s c t at
len (unbondnl) El ïmmcementls El 3mm w/ Metal tMdced onlay
El 2 mm with keel
Fig. 3 (a) Maximum and minimum principai strains for center and offset ioading in the anterior-posterior direction, (b) Contour plots of maximum and minimum principai strains at the bone interface for cemented components centraily ioaded. Journal of Biomechanical Engineering
Fig. 5 Contour piots of stresses and strains at the bone interface surface of an Ei 3 mm component and El 2 mm component with keel design JUNE2014, Vol. 136 / 061008-3
Discussion In this study, we presented a novel approach for the treatment of early OA called the early intervention design which reverses the material scheme between femoral and tibial components compared with traditional UKA components. Such a scheme was introduced by Charnley [28] in the mid 1970s but was limited in use. Using today's materials and techniques, the concept may well be feasible. With a reversed materials scheme, the El design requires minimal tibial bone resection which preserves the dense and stiff bone near the surface of the proximal tibia. FEA methods were used to investigate the stress, strain, and SED differences between El tibial components, an intact tibia model, and traditional UKA tibial components. The all-plastic inlay showed elevated stresses, strains, and SED compared with an intact reference model. El 2 mm. El 3 mm, and a metal-backed onlay. For the inlay, the elevated SED, stress, and strain may be indicators of loosening, bone remodeling, and pain reported clinically [26,29-31] caused in part by the low modulus of the plastic and its thickness. Although FE simulations used idealized assumptions, such as even cement layers and complete bonded contact between surfaces, an all-plastic inlay component produced very high strains which may explain its tendency to fail early as shown in published experimental and clinical studies. Experimental studies have reported higher strains associated with inlay designs compared with metal-backed onlays using implanted tibia models [31,32], while clinically, inlays have been shown to fail much sooner than metal-backed onlay designs [33,34]. The causes of tibial component subsidence and loosening are still largely unknown; however, it is believed that a number of factors, such as bone strength and overall design of a tibial component, can lead to subsidence. These two factors also affect strains within the supporting bone as demonstrated in this study. For an El 2 mm, the small thickness most likely caused the elevated stresses, strains, and SED. A cement layer helped to reduce stresses and strains in the supporting bone [35]. Similarly, with the addition of a central keel, the El 2 mm showed reduced stress and strain values when compared with an El 3 mm and metalbacked onlay. The keel distributed the load more effectively onto the underlying bone by stiffening the center of the component where the load is applied. Additionally, dishing the upper surface of the El design further stiffened the component. In terms of preserving strong bone, low resection depths are an advantage. The bone strength and density have been found to diminish with depth below the surface, in both normal [36] and osteoarthritic cases [37,38]. An El 3 mm and metal-backed onlay produced comparable stresses and strains when loaded centrally, anteriorly, and posteriorly; however, an El 3 mm component only requires a 2 mm resection, whereas a metal-backed onlay requires a 6-8 mm resection. With less resection, the stronger and denser bone would be preserved providing stronger fixation of the component and ease of a future revision surgery since more bone stock is available. In this study, we used a typical arthritic tibia with a nonuniform distribution of bone properties in the proximal tibia. The importance of using such a bone model was that it affected the stress and strain distributions and took account of the levels of resection required for each type of component. Additional evaluation parameters will now be necessary to test the viability of the concept in the tibia. From a practical point of view, in cemented components, cement penetration will be restricted in dense bone, and hence experiments will be necessary using techniques, such as microdrilling. We analyzed one tibial sample, but in OA, the bone density patterns vary widely [37,38], requiring analyses which cover the range. Axial loads were applied centrally, anteriorly, and posteriorly to all models analyzed here, however, other force components representing a range of loading scenarios need to be further investigated. This is particularly important if El components are to be 061008-4 / Vol. 136, JUNE 2014
used uncemented where interface stability is essential for osseointegration, although it should be noted that for offset loading in the bonded condition, there were minimal tensile stresses found across the interface. For the cementless El components, minimal differences in stress, strain, or SED were found between bonded and unbonded cases. For cementless El components, it will be desirable to have a very rough underside for initial stability [10]. The femoral component also needs analysis, using, for example, the configuration of Charnley, a 6-8 mm plastic inlay embedded in the distal femur, consistent with the osteoarthritic lesion in early OA [39]. This component could be all-pla.stic or have a metal backing. In either case, even if it had to be revised to a total knee in the future, this would involve 10 mm of distal femoral resection, completely containing the implant. In summary, FE simulations detnonstrated the effectiveness of using an early intervention tibial component over traditional UKA tibial components. The El design will require small incisions and bone resections, which will allow for natural anatomy of the patient to be preserved and less invasive surgery techniques to be used. This study showed that an El design has the potential to reduce the loosening rates seen in contemporary UKA designs.
Acknowledgment The Department of Orthopaedic Surgery, New York University, Hospital for Joint Diseases funded this study. We would like to thank Professor Nikhil Gupta and Dr. Oran Kennedy for helpful discussions. We are also grateful to Dr. Joseph Bosco and Dr. Ivan Fernandez-Madrid from New York University, Hospital for Joint Diseases for their valuable medical perspective regarding component feasibility. This study was selected for the ISTA 2013 Student Biomechanics Award.
References [1] Insall, J., and Aglietti, P., 1980, "A Five to Seven-Year Follow-Up of Unicondylar Arthroplasty," J. Bone Joint Surg. Am., 62, pp. 1329-1337. [2] Swank, M., Stulberg, S. D., Jiganti, J.. and Machairas, S.. 1993. "The Natural History of Unicompartmental Arthroplasty. An Eight-Year Follow-Up Study With Survivorship Analysis," Clin. Orthop. Relat. Res., 286, pp. 130-142. [3] Cheng, T., Guoyou, Z., and Zhang, X., 2010, "Does Minimally Invasive Surgery Improve Short-Term Recovery in Total Knee Arthroplasty?," Clin. Orthop. Relat. Res., 468, pp. 1635-1648. |4] McAllister, C , and Stepanian, J., 2008, "The Impact of Minimally Invasive Surgical Techniques on Early Range of Motion after Primary Total Knee Arthroplasty," J. Arthroplasty, 23(1), pp. 10-18. 15] Khanna, A., Gougoulias, N., Longo U. G., and Maffulli, N., 2009, "Minimally Invasive Total Knee Arthroplasty: A Systematic Review." Orthop. Clin. North Am., 40(4), pp. 479^89. |6] Borus, T., and ThornhiU, T., 2008, "Unicompartmental Knee Arthroplasty," J. Am. Acad. Orthop. Surg., 16(1). pp. 9-18. [71 Tanavalee, A., Choi, Y. J., and Trai, A. J., 2005, "Unicondylar Knee Arthroplasty: Past and Present," Orthopedics, 28, pp. 1423-1433. [8) Rho, J. Y., Hobatho, M. C , and Ashman, R. B., 1995. "Relations of Mechanical Properties to Density and CT Numbers in Human Bone," Med. Eng. Phys., 17(5), pp. 347-355. [91 Gray, H., Taddei, F., Zavatsky, A., Cristofolini, L., and Gill, H., 2008, "Experimental Validation of a Finite Element Model of a Human Cadaveric Tibia," ASME J. Biomech. Eng., 130, p. 031016. [10] Zhang, Y., Ahn, P. B., Fitzpatrick. D. C, Heiner, A., Poggie, R. A., and Brown, T. D., 1999, "Interfacial Frictional Behavior; Cancellous Bone, Cortical Bone, and a Novel Porous Tantalum Biomaterial," J. Musculoskeletal Res., 3(4), pp. 245—251. [11] Tandon. R., 1999, "Net Shaping of Co-Cr-Mo (F-75) via Metal Injection Molding," Cohall-Base Alloys for Biomédical Applications, J. A. Disegi, R. L. Kennedy, and R. Pilliar, eds., ASTM International, West Conshohocken, PA, pp. 3-10. [12) Rapperport, D. J., Carter, D. R., and Schurman, D. J., 1987, "Contact Finite Element Stress Analysis of Porous Ingrowth Aeetabular Cup Implantation, Ingrowth, and Loosening," J. Orthop. Res., 5(4), pp. 548-561. [13] Shanbhag, A. S., Hasselman, C. T., and Rubash, H. E., 1996, "Technique for Generating Submicrometer Ultra High Molecular Weight Polyethylene Particles," J. Orthop. Res., 14, pp. 1000-1004. [14] Kurtz, S. M.. Villarraga, M. L.. Herr, M. P., Bergström, J. S., and Rimnac, C. M., 2002, "Thermomechanical Behavior of Virgin and Highly Crosslinked Ultra-High Molecular Weight Polyethylene Used in Total Joint Replacements," Biomaterials, 23(17), pp. 3681-3697. [15] Ho, S. P., Carpick. R., Boland, T., and LaBerge, M., 2002, "Nanotribology of Cocr-UHMWPF TJR Prosthesis Using Atomic Force Microscopy," Wear, 253(11-12), pp. 1145-1155.
Transactions of the ASME
[16] Lakes, R. S., 2002, Composite Biomaterials. Biomaterials: Principles and Applications, CRC Press, LLC, Boca Raton, FL, p. 84. [17] Orr, J. F., Dunne, N. J., and Quinn, Q. C, 2003, "Shrinkage Stresses in Bone Cement," Biomaterials, 24(17), pp. 2933-2940. [18] Mündemiann, A., Dyrby, C. O., D'Lima, D. D.. Colwell, C. W., Jr., and Andriacchi, T. P., 2008, "in vivo Knee Loading Characteristics During Activities of Daily Living as Measured by an Instrumented Total Knee Replacement," J. Orthop. Res., 26(9), pp. 1167-1172. [19] Argenson, J. A., Komistek, R. D., Aubaniac, J., Dennis, D. A., Northcut, E., Anderson, D., and Agostini, S., 2002, "In Vivo Determination of Knee Kinematics for Subjects Implanted With a Unicompartmental Arthroplasty," J. Arthroplasty, 17(8), pp. 1049-1054. [20] Banks, S., Fregly, B., Boniforti, F., Reinschmidt, C, and Romagnoli, S., 2005, "Comparing In Vivo Kinematics of Unicondylar and Bi-Unicondylar Knee Replacements," Knee, 13, pp. 55Í-556. [21] D'Lima, D. D., Patil, S., Steklov, N., and Colwell, C. W., 2011, "The 2011 ABJS Nicolas Andry Award: 'Lab'-In-A-Knee: in vivo Knee Forces, Kinematics, and Contact Analysis," Clin. Orthop. Relat. Res., 469(10), pp. 2953-2970. [22] Fukubayashi, T.. and Kurosawa, H., 1980, "The Contact Area and Distribution Pattem of the Knee," Acta Orthop. Scand., 51, pp. 871-879. [23] Bell, C , Amo, S., Hadley, S., Campbell, K., Beltran L., Recht M., Sherman O., Walker P., 2012. "Relative Contacts Between the Medial Meniscus and Exposed Cartilage in the Knee," 58th Annual Orthopaedic Research Society. [24] Taylor, M., Tanner, K. E., and Freeman, M. A. R., 1998, "Finite Element Analysis of the Implanted Proximal Tibia: A Relationship Between the Initial Cancellous Bone Stresses and Implant Migration," J. Biomech., 31, pp. 303-310. [25] Huiskes, R., Weinans, H., Grootenboer, H. J., Dalstra, M., Fudala, B., and Sloof, T. J., 1987, "Adaptive Bone-Remodeling Theory Applied to ProstheticDesign Analysis," J. Biomech., 20(11-12), pp. 1135-1150. [26] Simpson. D., Kendrick, B., Dodd, C, Price, A., Gill, H., and Murray, D., 2010, "Load Transfer in the Proximal Tibia Following Implantation With a Unicompartmental Knee Replacement: A Static Snapshot," Proc. Inst. Mech. Eng., Part H, 225, pp. 521-529. [27] Morgan, E. F., and Keaveny, T. M., 2001, "Dependence of Yield Strain on Human Trabecular Bone on Anatomic Site," J. Biomech., 34, pp. 569—577.
Journal of Biomechanical Engineering
[28] Minas, R. J., Day, J. B., and Hardinge, K., 1982, "Kinesiologic and Biomechanical Assessment of the Chamley 'Load Angle Inlay' Knee Prosthesis," Eng. Med., 11(1), pp. 25-32. [29] Arastu, M., Vijayaraghavan, J., Chissell, H., Hull, J., Newman, J., and Robison, J., 2009, "Early Failure of a Mobile Bearing Unicompartmental Knee Replacement," Knee Surg. Sports Traumatol. Arthrose, 17(10), pp. 1178-1183. [30] Lisowski L. A., van den Bekerom M. P. J., Pilot P., van Dijk C. N., Lisowski A. E., 2011, "Oxford Phase 3 Unicompartmental Knee Arthroplasty: MediumTerm Results of a Minimally Invasive Surgical Procedure," Knee Surg Sports Traumatol Arthrose, 19(2): 277-84. [31] Small, S. R., Berend, M. E., Ritter, M. D., Buckley, C. A., and Rogge, R. D., 2011, "Metal Backing Significantly Decreases Tibial Strains in a Medial Unicompartmental Knee Arthropiasty Model," J. Arthroplasty, 26(5), pp. 777-782. [32] Scott, C. E. H., Eaton, M. J., Nutton, R. W.. Wade, F., Pankaj, P., and Evans, S., 2013, "Proximal Tibial Strain in Medial Unicompartmental Knee Replacements," Bone Joint J., 95B, pp. 1339-1347. [33] Saenz, C. L., McGrath, M. S., Marker, D. R., Seyler, T. M., Mont, M. A., and Bonutti, P. M., 2010, "Early Failure of a Unicompartmental Knee Arthroplasty Design With an All-Polyethylene Tibial Component," Knee, 17, pp. 53-56. [34] Gladnick, B., Nam, D., Khamaisy, S., Paul, S., and Pearle. A., "Inlay Versus Onlay Tibial Implants in Robotic Unicondylar Knee Arthroplasty," Annual Congress of the Intemational Society for Technology in Arthroplasty, Palm Beach, FUOct. 16-19. [35] Thompson, M. T., Conditt, M. A., Otto, J. K., and Redish, M., 2010, "The knportance of Good Cement Mantle With an All-Poly Inlay UKA" Transactions of the 56th Annual Meeting—Orthopaedic Research Society, Orthopaedic Re.search Society. [36] Hvid, 1., and Hansen, S. L., 1985, "Trabecular Bone Strength Pattems at the Proxitnal Tibial Epiphysis," J. Orthop. Res., 3(4), pp. 464-472. [37] Hvid, I., Í988, "Trabecuiar Bone Strength at the Knee," Clin. Orthop. Relat. Res., 227, pp. 210-221. [38] Wong, N., Wei, C. S.. Gautam, P., and Walker, P. S., 2012, "Multi-Planar Visualization of Tibial Bone Density Distribution From a Tibial CAT Scan Dataset," Orthopedics Research Society Annual Meeting, February, 2012. [39] Arno, S., Maffei, D., Walker, P. S., Schwarzkopf, R., Desai, P., and Steiner, G., 2011, "Retrospective Analysis of Total Knee Arthroplasty Cases for Visual, Histological and Clinical Eligibility of Unicompartmental Knee Arthroplasties," J. Arthroplasty, 26(8), pp. 1396-1403.
JUNE2014, Vol. 136 / 061008-5
Copyright of Journal of Biomechanical Engineering is the property of American Society of Mechanical Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
