Total Knees Designed for Normal Kinematics Evaluated in an Up-and-Down Crouching Machine

Transcription

1 Total Knees Designed for Normal Kinematics Evaluated in an Up-and-Down Crouching Machine Gokce Yildirim, Peter S. Walker, Jason Boyer Department of Orthopaedic Surgery, NYU Hospital for Joint Diseases, c/o Veterans Administration Medical Center, 423 East 23rd Street, Annex Building #2, Room 206A, New York, New York Received 9 January 2008; accepted 20 November 2008 Published online 15 January 2009 in Wiley InterScience ( DOI /jor ABSTRACT: We constructed a crouching machine to study the motion of the knee joint, in which a motor was used to wind the quadriceps tendon so as to move the knee from high flexion to extension and back into flexion, while springs simulated hamstrings forces. Seven human cadaveric knees were tested intact and then after anterior cruciate ligament (ACL) resection. Motions of the femur, tibia, and patella were recorded by an optical tracking system. We then inserted plastic models representing commonly used total condylar and posterior stabilized knee replacement designs. Femoral motion was described by successive positions of the transverse axis of the femur projected onto the tibial surface. In the knee replacements, motions were similar to that of an ACL-deficient knee. We then tested two new designs with features intended to prevent anterior paradoxical sliding and to promote a medial pivot motion with femoral rollback primarily on the lateral side. The motion path more closely followed that of the normal intact knee. We concluded that motion guiding features in a total knee replacement could reproduce a normal neutral path that might result in functional improvements for the patient. ß 2009 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 27: , 2009 Keywords: total knee design; total knee kinematics The neutral path of motion of the knee in both loaded and unloaded conditions has been determined using cadaveric specimens or living subjects. 1 9 Main findings have been posterior displacement of the lateral femoral condyle with flexion and a relatively immobile medial femoral condyle, producing internal tibial rotation of about 208 through a full range of flexion. Likely advantages of posterior displacement include an increased quadriceps lever arm and the facilitation of a high flexion angle by delaying impingement of the posterior tibia with the posterior femoral cortex Minimal anteroposterior (AP) displacement of the medial condyle protects the integrity of the meniscus, 13 while abnormal tibial rotation in flexion changes the relative medial to lateral facet forces of the patellofemoral joint. 14 A medial side that is stable in the AP direction might be important for achieving normal function. 15 After total knee replacement (TKR), however, the kinematics are abnormal. Paradoxical anterior femoral sliding with increasing flexion was found in many studies, while posterior displacement was found in high flexion in posterior stabilized (PS) knees, but less consistently in cruciate retaining (CR) knees Internal tibial rotation with flexion averaged only a few degrees for both CR and PS TKRs, as well as in mobile bearing knees. 20 Similar to the case for the normal knee, the advantages of posterior femoral displacement with flexion, and preserving the posterior condylar offset, was demonstrated. 18,21,22 In these studies, however, considerable variation was found in motion paths among activities and individuals. We hypothesized that a normal neutral path of motion could be restored with a TKR by adding motion-guiding Correspondence to: Peter S. Walker (T: , x6444; F: ; ptrswlkr@aol.com) ß 2009 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. features designed to preserve a stable medial side, produce lateral femoral rollback, and prevent anterior femoral sliding in early flexion. We studied this hypothesis by testing intact human cadaveric knees and then again after implanting models of different TKR designs in a dynamic knee testing machine. MATERIALS AND METHODS Four TKR types were designed using Rhinoceros 3.0 Nurbs software (Mc Neel & Associated, Seatlle, WA) (Fig. 1): 1. Total Condylar Prosthesis (TCP): Standard total condylar replacement with partially conforming doubledished surfaces with Anterior Cruciate Ligament (ACL) and Posterior Cruciate Ligament (PCL) resection. 2. PS: Standard PS with the same geometry as the TCP but with a central post-cam for femoral rollback after 758 flexion. 3. Experimental Design 1 (EXP1). A modified PS for which the medial tibial surface was more dished and the lateral surface less dished to provide medial stability and promote external femoral rotation about the tibial axis with flexion (equivalent to internal tibial rotation). In addition, an anteromedial recess-ramp feature was added to restrict anterior femoral displacement in the first half of the flexion range. 4. Experimental Design 2 (EXP2). A surface-guided design for which the depth of the femoral intercondylar groove was reduced from 8 mm at the distal end of the femur to 0 mm at the posterior of the femur. This groove interfaced with a central tibial ramp to produce posterior femoral displacement with flexion. The medial and lateral tibial dishing and the anteromedial recess-ramp were as in EXP1. All of the TKR models were made from stereolithography (stl) computer files, using rigid polyurethane material (U850, 75 shore D hardness, lubricated coefficient of friction of 0.03). A lower extremity test machine was built (Fig. 2) to replicate a crouching up-and-down motion. The tibia was attached to a fixture through an intramedullary rod that was free to rotate 1022

2 TOTAL KNEE DESIGNED FOR NORMAL KINEMATICS 1023 Figure 2. The crouching up-and-down machine, which applied continuous flexion extension to the knee specimens. Figure 1. Anteromedial (A) and posterolateral (B) views of the four TKR designs used in the experiments. about its long axis. The distal tibia was hinged at the ankle joint, but the ankle was attached to the machine s frame by flexible polyurethane blocks to allow for a few degrees of varus valgus rotations as in everyday activities. 23 The hip joint was simulated with a spherical bearing attached to a plate that was allowed to slide vertically along a smooth rail so the knee could flex to The total vertical force at the hip was 73.4N. A stepper motor was housed in a sub-assembly attached to the femoral fixture. The motor activated a pulley and force transducer, which were attached to the quadriceps tendon through a clamp grasping a looped cloth sutured to the tendon. The mean quadriceps force was 44 N at 158 flexion, reaching a maximum of 267 N at 908. To simulate the hamstring muscles, two extension springs were screwed to the posterior tibia at the hamstring attachment locations and attached to the femoral fixture. The forces were an average of 172 N with the knee at 158 flexion (86 N in each spring), reducing to a total of 44N with the knee at This pattern of reducing hamstrings activity with flexion was based on electromyography (EMG) data. 24 To track the location of the knee during the tests, an optical motion tracking system (MicronTracker, Claron Technology, Toronto, Canada) with 0.2 mm accuracy followed targets pinned to each of the three bones. Three conical holes (1 mm diameter, two holes medially, one lateral) were made in the distal end of the femoral cortex and the proximal end of the tibial cortex to act as fiducial points. A digitizer (MicroScribe, Immersion Corp., San Jose, CA) with 0.23 mm accuracy was used to record the fiducial points with respect to the marker positions before the start of the test. Seven fresh frozen human cadaveric knee joints were tested. The intact knee was first tested for three cycles of flexion extension, with a 1 2 s pause at each extreme. The range of flexion was nominally 15 to Because some knees had an extension lag and because knee motion was difficult to control with the quadriceps close to extension, 158 was used rather than 08. The ACL was inspected visually and by palpation by the surgeon (J. B.); it was then transected, and the test was repeated. The PCL and the menisci were then removed. All articular cartilage surfaces were digitized with a point spacing of about 3 mm, along with the corresponding fiducial points. After the digitization, TKR was performed by the surgeon using standard instruments, based on the NexGen PS system (Zimmer Inc., Warsaw, IN). Posterior femoral condylar referencing was used to define 38 of femoral component rotation. The AP tibial axis was defined by a line joining the center of the PCL attachment to the medial border of the patellar tendon. Each of the four TKR models was inserted in turn in the knee using a removable putty to hold the components in place, and the test was repeated after digitizing the fiducial points on the TKR. Lubrication was provided by a low-friction gel (Krytox TA grease, Miller-Stephenson Chemical Co., Danbury, CT), which in a separate test was found to give a friction coefficient of 0.03, somewhat lower than that of metal on polyethylene. For each test, RapidForm (Inus Technologies, Sunnyvale, CA) surfacing software was used to recreate the positions of the knee in 158 increments. To describe motion of the intact knee, two equal circles (average 23 mm radius) were visually fit 46 mm apart on the lateral and medial femoral condyles, on the posterior section of the articular cartilage (Fig. 3). The centers were connected with a vector to define the transverse femoral axis, corresponding to a posterior cylindrical or spherical axis rather than an epicondylar axis. 25 The femoral transverse axis was then projected on to the upper tibial surface, and the

3 1024 YILDIRIM ET AL. Figure 3. Definition of the transverse femoral axis, CL CM. The axis was projected onto the proximal tibia, TL TM, at a succession of flexion angles to describe the motion. succession of these axes for the 158 angular increments represented the motion path of the femur on the tibia. The tibial transverse axis was defined as being parallel to the line across the posterior edge of the tibial component. The maximum displacements of the lateral and medial sides of the femoral transverse axis and the associated internal external rotation angles were calculated for the full flexion range. To compare the intact motions with those after TKR, the tibial transverse axis was defined to be the same in both cases. Statistical differences in the displacements of the knee between intact, ACL-deficient, and the four TKRs were tested using paired t-test analysis. RESULTS The motion paths of the transverse femoral axis were the same when the knee was flexing and extending (Fig. 4). For the intact knee, much more posterior displacement of the lateral femoral condyle was found compared with the medial side, producing external femoral rotation with flexion. This fan pattern was reproduced with TKR designs EXP1 and EXP2. Figure 4. A representative example of the paths of the transverse femoral axis for the full range of flexion for the intact knee ACL (þ), the ACL-deficient knee ACL ( ), and for the four TKRs. For the ACL-deficient knee, the transverse femoral axis fluctuated in rotation and position throughout flexion, but gave more similar displacements on the lateral and medial sides, resulting in relatively lower values of femoral rotation. A similar pattern was seen for the TCP and PS designs; the posterior translation of the PS was greater beyond 758 due to cam-post engagement. Displacement and rotation results for all seven knees from flexion are shown in Table 1. In two knees, the ACLs were thin and appeared to be nonfunctional, so the intact data were not included. In a third knee, the ACL was somewhat compromised, but the intact data were included. Our null hypothesis was that the displacements and rotations were different between our test groups. The intact and ACL-deficient conditions showed significant differences for lateral displacement (p ¼ 0.04) and for rotation (p ¼ 0.02). Significant differences were also found for medial displacement and rotation between the intact knee and the TCP and PS TKRs (p < 0.02). The only significant differences between intact and the experimental designs were for medial displacement in EXP1 (p ¼ 0.001) and rotation for EXP2 (p ¼ 0.007). Comparing the ACL-deficient knee with the TCP and PS, significant differences were found in rotation and medial displacement (p < 0.03). When the ACL-deficient knee was compared with EXP1 and EXP2, significant differences were found in lateral displacement and rotation (p < 0.01), but no significant difference occurred for the medial displacement of EXP1 and EXP2. Regarding lateral and medial displacements, there were significant differences (p < ) for the intact knee, and for EXP1 and EXP2, but not for the ACL-deficient knee and the PS. DISCUSSION Our overall conclusions were that the EXP1 and EXP2 designs behaved more like an intact knee, whereas the TCP and PS resembled more an ACL-deficient knee. The notable kinematic features were the larger lateral than medial displacements, producing axial rotation, a characteristic of the normal intact knee. Our test machine is a derivative of the Oxford rig, 26 which has been used for studying the mechanics of the knee under simulated muscle activity. The inclusion of the hamstrings is important in terms of modifying the shear force and hence the AP position of the femur on the tibia The provision of varus valgus rotation at the ankle was indicated in a study of crouching in high flexion. 23 The reduced forces in our tests represent about 20% of actual forces, based on in vivo data from an instrumented TKR. 30 High flexion activities are performed less frequently than walking and with the most difficulty, but are important to many patients. 31 The motions of the normal intact specimens were consistent with previous studies (cited in the Introduction). After ACL resection, a significant change in kinematics occurred, notably a reduction in the rotation due to the reduced posterior displacement of the lateral

4 TOTAL KNEE DESIGNED FOR NORMAL KINEMATICS 1025 Table 1. Total AP Displacements on the Lateral and Medial Sides, and the Axial Rotations for the Flexion Range of for Seven Knees a Lateral A/P Medial A/P Rotation (deg) Lateral A/P Medial A/P Rotation (deg) INTACT NO ACL Avg Avg SD SD EXP1 TCP Avg Avg SD SD EXP2 PS Avg Avg SD SD a Data are for the intact knee, after ACL resection, and after insertion of the four test TKRs. femoral condyle and increased displacement on the medial side. In addition, motion tended to be erratic. This behavior is consistent with ACL anatomy, especially the antero-medial band, 32 which tightens in flexion. 33 Reuben et al. 34 showed instability in AP tibial translations after ACL resection, up to about 608 flexion. In vivo studies showed abnormalities in rotation and translations in various activities after ACL injury, 35 although hamstrings function could alleviate this in stair ascent. 36 All of this evidence indicates the importance of substituting for ACL function in TKR, if normal kinematics are to be achieved. Reproducing normal stability and kinematics in a TKR has been approached in several ways. 37 The original TCP concept provided for stability and laxity, but did not provide rollback in flexion. 38,39 Posterior rollback was achieved in TCP designs by reducing posterior femoraltibial conformity and preserving the PCL (CR designs). Rollback, but not rotation, was also achieved by the use of an intercondylar cam-post to replace PCL function, 40,41 a concept widely used today in the PS designs. But all of these condylar designs provide AP and rotational stability to some extent by the amount of conformity between the femoral and tibial surfaces. The concept of providing AP stability only on the medial side, but allowing rollback and rotation to occur on a less-dished lateral side was first employed in a variation of the Freeman-Samuelson design. 42 This concept was further developed with a low-conformity lateral side and PCL retention, the so-called medial pivot design. 43 During certain tests, this did achieve lateral femoral translation up to mid-flexion and avoided paradoxical motion. 43 Similar tibial geometry, but using a central post-cam instead of preserving the PCL, also demonstrated more anatomic motion characteristics, both in a computer model and in patients. 44,45 Computer-generated designs were proposed that would guide both rollback rotations by converging bearing surfaces, to some extent replicating anatomic features. 46 In the designs in our study, we attempted to reproduce normal kinematics by making the medial side stable, but with some partial conformity posteriorly to allow for posterior femoral displacement in high flexion. To limit anterior femoral displacement in early flexion, a medial femoral recess-tibial ramp arrangement was designed, where the recess was a continuation of the distal-posterior sagittal radius. This medial-side stability would effectively reproduce the function of both cruciate ligaments. However, to simulate PCL function in higher flexion, a central cam-post mechanism was used. In the central post design, the curved contours of the post would likely prevent polyethylene damage from occurring, a phenomenon that has been reported for many current PS designs The low-profile tibial ramp EXP1 had the advantage that a full-contact patellar bearing surface was preserved on the femoral component up to full flexion. Dislocation per se could not occur with this design, although with high shear loading in high flexion there could be some sliding up the ramp. When inserting designs modeled on EXP1 and EXP2 with a relatively high medial conformity, it may be important to avoid potential excessive constraint by providing for laxity on the medial side to avoid possible conflict with the medial collateral ligament (MCL) and other structures. As for wear, a more conforming medial side that limits AP sliding on the more highly loaded side of the joint may reduce wear, due to the lower contact stresses, as would a more controlled motion path compared to a more erratic path as has been described for current TKR designs in fluoroscopic studies. Furthermore, notwithstanding the variations that occur in arthritic knees, a normal motion path is likely more compatible with soft tissues and muscles than an abnormal path. Adequate AP and rotational laxity is inherent in the experimental designs, primarily on the lateral side, which allows for variability of femoral and tibial component positioning in extension, as well as variability in motion paths in different activities. In conclusion, we showed that certain design configurations used widely today do not restore normal kinematics, even though the clinical results with these implants have generally been successful. In contrast, by adding certain geometrical features in a guidedmotion knee, more normal motion patterns were restored. Future clinical studies with such designs

5 1026 YILDIRIM ET AL. will be required to determine if there are functional advantages as a result. ACKNOWLEDGMENTS The test machine was constructed by Mr. Daniel Hennessy. The work was funded in part by a grant from Zimmer, Inc. (Warsaw, IN). Statistical advice was given by Dr. Michael Waigh. REFERENCES 1. Kurosawa H, Walker PS, Abe S, et al Geometry and motion of the knee for implant and orthotic design. J Biomech 18: Iwaki H, Pinskerova V, Freeman MAR Tibiofemoral movement 1: the shapes and relative movements of the femur and tibia in the unloaded cadaver knee: studied by dissection and MRI. J Bone Joint Surg [Am] 82B: Nakagawa S, Kadoya Y, Todo S, et al Tibiofemoral movement 3: full flexion in the living knee studied by MRI. J Bone Joint Surg [Br] 6: Hollister AM, Jatana S, Singh AK, et al The axes of rotation of the knee. Clin Orthop 290: Rovick JS, Reuben JD, Schrager RJ, et al Relation between knee motion and ligament length patterns. Clin Biomech 6: Patel VV, Hall K, Ries M, et al A three-dimensional MRI analysis of knee kinematics. J Orthop Res 22: Williams A, Logan M Understanding tibio-femoral motion. Knee 11: Hill PF, Vedi V, Williams A, et al Tibiofemoral movement 2: the loaded and unloaded living knee studied by MRI. J Bone Joint Surg [Br] 82: Komistek RD, Dennis DA, Mahfouz M In vivo fluoroscopic analysis of the normal human knee. Clin Orthop 410: Li G, Zayontz S, DeFrate LE, et al Kinematics of the knee at high flexion angles: an in vitro investigation. J Orthop Res 22: Conditt MA, Thompson MT, Wenk TJ, et al Knee kinematics and medial lift-off during high flexion activities. Proceedings of the 52nd Annual Orthopaedic Research Society, Vol 31, Paper 0251, March 18 22, Chicago, Illinois. 12. Yildirim G, Walker PS, Sussman-Fort J, et al The contact locations in the knee during high flexion. Knee 14: Von Eisenhart-Rothe R, Bringmann C, Siebert M, et al Femoro-tibial and menisco-tibial translation patterns in patients with unilateral anterior cruciate ligament deficiency a potential cause of secondary meniscal tears. J Orthop Res 22: Hefzy MS, Jackson WT, Saddemi SR, et al Effects of tibial rotations on patellar tracking and patello-femoral contact areas. J Biomed Eng 14: Blaha JD The rationale for a total knee implant that confers anteroposterior stability throughout range of motion. J Arthroplasty 19(Suppl 1): Dennis D, Komistek RD, Mahfouz MR, et al Multicenter determination of in vivo kinematics after total knee arthroplasty. Clin Orthop 416: Stiehl J, Komistek RD, Cloutier JM, et al The cruciate ligaments in total knee arthroplasty. J Arthroplasy 15: Banks SA, Harman MJK, Bellemans J, et al Making sense of knee arthroplasty kinematics: news you can use. J Bone Joint Surg [Am] 85: Dennis DA, Komistek RD, Mahfouz MR, et al A multicenter analysis of axial femorotibial rotation after total knee arthroplasty. Clin Orthop 428: Delport HP, Banks SA, De Schepper J, et al A kinematic comparison of fixed- and mobile-bearing knee replacements. J Bone Joint Surg [Br] 88B: Bellemans J, Banks S, Victor J, et al Fluoroscopic analysis of the kinematics of deep flexion in total knee arthroplasty: influence of posterior condylar offset. J Bone Joint Surg [Br] 84B: Walker PS, Yildirim G, Sussman-Fort J, et al Factors affecting the impingement angle of fixed- and mobile-bearing total knee replacements a laboratory study. J Arthroplasty 22: Hemmerich A, Brown H, Smith S, et al Hip, knee, and ankle kinematics of high range of motion activities of daily living. J Ortho Res 24: Kawagoe S, Tajima N, Chosa E Biomechanical analysis of effects of foot placement with varying chair height on the motion of standing up. J Orthop Sci 5: Eckhoff D, Hogan C, DeMatteo L, et al Difference between the epicondylar and cylindrical axis of the knee. Clin Orthop 461: Biden E, O Connor J Experimental methods used to evaluate knee ligament function. In: Daniel D, Akeson WH, O Connor J, et al. editors. Knee ligaments: structure, function, injury, and repair. New York: Raven Press; p MacWilliams BA, Wilson DR, DesJardins JD, et al Hamstrings cocontraction reduces internal rotation, anterior translation, and anterior cruciate ligament load in weightbearing flexion. J Orthop Res 17: Most E, Li G, Park SE, et al Tibiofemoral contact after conventional and high-flexion posterior cruciate-retaining total knee arthroplasty. Proceedings of the 50th Annual Meeting of ORS, poster No. 1397, March 7 10, San Francisco, California. 29. Steele JR, Roger GJ, Milburn PD Tibial translation and hamstring activity during active and passive arthrometric assessment of knee laxity. Knee 4: D Lima DD, Patil S, Steklov N, et al In vivo knee moments and shear after total knee arthroplasty. J Biomech 40:S11 S Weiss JM, Noble PC, Conditt MA, et al What functional activities are important to patients with knee replacements? Clin Orthop 404: Girgis FG, Marshall JL, Monajem A The cruciate ligaments of the knee joint: anatomical functional and experimental analysis. Clin Orthop 106: Amis AA, Dawkins GPC Functional anatomy of the anterior cruciate ligament: fibre bundle actions related to ligament replacements and injuries. J Bone Joint Surg [Br] 73B: Reuben JD, Rovick JS, Schrager RJ, et al Threedimensional dynamic motion analysis of the anterior cruciate ligament deficient knee joint. Am J Sport Med 17: DeFrate LE, Papannagari R, Gill TJ, et al The 6 degrees of freedom kinematics of the knee after anterior cruciate ligament deficiency: an in vivo imaging analysis. Am J Sport Med 34: Jonsson H, Kaarholm J Three-dimensional knee joint movements during a step-up: evaluation after anterior cruciate ligament rupture. J Orthop Res 12: Robinson RP The early innovators of today s resurfacing condylar knees. J Arthroplasty 20:2 26.