Contact stress at the post-cam mechanism in posterior-stabilised total knee arthroplasty

Transcription

1 Contact stress at the post-cam mechanism in posterior-stabilised total knee arthroplasty K. Nakayama, S. Matsuda, H. Miura, H. Higaki, K. Otsuka, Y. Iwamoto From Kyushu University, Fukuoka, Japan We measured the contact areas and contact stresses at the post-cam mechanism of a posterior-stabilised total knee arthroplasty when a posterior force of 500 N was applied to the Kirschner Performance, Scorpio Superflex, NexGen LPS Flex Fixed, and NexGen LPS Flex Mobile knee systems. Measurements were made at 90, 120, and 150 of flexion both in neutral rotation and 10 of internal rotation of the tibial component. Peak contact stresses at 90, 120, and 150 were 24.0, 33.9, and 28.8 MPa, respectively, for the Kirschner; 26.0, 32.4, and 22.1 MPa, respectively, for the Scorpio; and 34.1, 31.5, and 32.5 MPa, respectively, for the NexGen LPS Flex Fixed. With an internally rotated tibia, the contact stress increased significantly with all the fixed-bearing arthroplasties but not with the NexGen LPS Flex Mobile arthroplasty. The post-cam design should be modified in order to provide a larger contact area whilst avoiding any impingement and edge loading. " K. Nakayama, MD, Orthopaedic Surgeon " S. Matsuda, MD, PhD, Assistant Professor, Orthopaedic Surgeon " H. Miura, MD, PhD, Associate Professor, Orthopaedic Surgeon " Y. Iwamoto, MD, PhD, Orthopaedic Surgeon, Professor, Chairman Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan. " H. Higaki, PhD, Professor, Mechanical Engineer " K. Otsuka, Mechanical Engineer Department of Mechanical Engineering, Kyushu Sangyo University, Shokadai, Higashi-ku, Fukuoka , Japan. Correspondence should be sent to Assistant Professor S. Matsuda; mazda@ ortho.med. kyushu-u.ac.jp 2005 British Editorial Society of Bone and Joint Surgery doi: / x.87b $2.00 J Bone Joint Surg [Br] 2005;87-B: Received 28 May 2004; Accepted after revision 1 October 2004 erior-stabilised total knee arthroplasty (TKA) was introduced in the mid 1970s. 1 The procedure was developed to prevent posterior subluxation of the tibia, as a replacement for a degenerative posterior cruciate ligament, and to improve the range of movement of the knee by allowing femoral rollback. It has been widely used for more than two decades and long-term follow-up studies have reported satisfactory results. 2-4 However, complications involving the post-cam mechanism, such as dislocation of the knee and fracture or severe wear of the post have been reported. 5-9 Recently, surgeons have sought a greater range of movement after TKA. 10 It is thought that patients who undergo a posterior-stabilised TKA will achieve a greater range of movement than those who undergo a cruciateretaining TKA because after a posterior-stabilised TKA posterior roll back of the femur is expected 11 and the posterior soft tissue is easily released. Some posterior-stabilised designs have been modified in order to reduce the complications which occur at full flexion of the knee. However, these modifications are mainly focused on the tibiofemoral articulation. Biomechanical studies of deep flexion of the knee have been performed and some have shown that a high anteroposterior shear force is generated at the tibiofemoral joint during deep flexion There is concern that a higher contact force would be applied to the post-cam mechanism with an increased angle of flexion after a posterior-stabilised TKA. It is important to know what level of contact stress is generated at the post-cam interface but, to date, this has not been investigated. This study was designed to assess the effects of the post-cam design on contact area, contact stress, and contact location at the post- and cam-mechanism of a posterior-stabilised arthroplasty. Many posterior-stabilised designs are available. In the sagittal plane, many of the cams are oval-shaped, but the direction of the apex of the cam differs. This study investigated three different designs, in which the apex of the cam was orientated distally, posteriorly, or posterodistally. We hypothesised that different directions of the apex would cause different changes in contact stress during flexion of the knee. In the axial plane, the post-cam configurations are grouped in one of two designs, flaton-flat or curved-on-curved. The design in the axial plane would significantly affect contact stress when the tibia is rotated. In this study, the effects of the post-cam design in the axial plane, the effects of mobile bearing, and the effects of the of the post relative to that of the cam were assessed when the tibial component was rotated. Materials and Methods The following posterior-stabilised TKAs were used: Kirschner Performance posterior-stabilised (Kirschner Medical Corporation, Timonium, Maryland, medium size); Scorpio VOL. 87-B, No. 4, APRIL

2 484 K. NAKAYAMA, S. MATSUDA, H. MIURA, H. HIGAKI, K. OTSUKA, Y. IWAMOTO Fig. 1a Fig. 1b The Kirschner arthroplasty. Figure 1a On the axial view, the post-cam mechanism has a curved configuration. There is little space between the tibial post and each femoral condyle (arrow). Figure 1b On the sagittal view, the apex of the cam is orientated posteriorly, and the cam has a relatively flat contact surface against the post in 90 of flexion (arrow). Fig. 3a Fig. 3b The NexGen Flex arthroplasty. Figure 3a On the axial view, the post-cam mechanism has a flat configuration. Figure 3b On the sagittal view, the apex of the cam has a dull curve and is orientated posterodistally. Fig. 2a Fig. 2b The Scorpio arthroplasty. Figure 2a On the axial view, the post-cam mechanism has a curved configuration. Figure 2b On the sagittal view, the apex of the cam is orientated distally, and the cam has a relatively flat contact surface posteriorly. Fig. 4 The parallel-axis six-link actuator. The femoral and tibial components were attached to the fixtures that were mounted in the actuator. Superflex (Stryker, Allendale, New Jersey, size 5), NexGen LPS Flex Fixed (Zimmer, Warsaw, Indiana, size D femoral component, size 4 tibial component); and NexGen LPS Flex Mobile (Zimmer, size D femoral component, size 4 tibial component) (Figs 1 to 3). The design characteristics and the dimensions of the post-cam mechanism are shown in Tables I and II. Each femoral component was attached to a fixture that provided a flexion-extension range from 90, 120, and 150. The tibial component was implanted in a block of metal in the neutral position. The femoral and tibial fixtures were mounted in a parallel-link six-axis actuator (Fig. 4). A compressive posterior load of 500 N, parallel to the tibial joint surface, was applied to the tibial component against the femoral component. The position of the femoral component in the sagittal plane was adjusted so that both medial and lateral femoral condyles contacted the tibial articular surfaces. A digital electronic stress sensor (K-Scan sensor, Tekscan Inc., Boston, Massachussetts) was placed at the post-cam interface to measure both contact stress and area. An IBM desktop computer connected to the sensor recorded the stress at each intersection at the point it was scanned. The accuracy of this entire system has been assessed previously. 18 Measurements were performed at 90, 120, and 150 of flexion using the Kirschner, the Scorpio, and the NexGen LPS Flex Fixed components with neutral rotation of the tibial component. Each measurement was performed five times for each component to allow calculation of the variance across the testing conditions. Peak contact stress, defined as the highest stress of all the sensing locations, mean contact stress, and contact area were automatically calculated by the Tekscan software. The centre of the contact area was also detected, and the distance from the deepest part of the upper tibial surface to the centre was measured (Fig. 5). Peak contact stress was then measured THE JOURNAL OF BONE AND JOINT SURGERY

3 CONTACT STRESS AT THE POST-CAM MECHANISM IN POSTERIOR-STABILISED TOTAL KNEE ARTHROPLASTY 485 Table I. Design characteristics of the post-cam of the four posterior stabilised TKAs Kirschner Scorpio NexGen LPS Flex Fixed NexGen LPS Flex Mobile Orientation of the cam apex erior Distal erodistal erodistal Configuration in the axial plane Curved-on-curved Curved-on-curved Flat-on-flat Flat-on-flat Mobility of the bearing Fixed Fixed Fixed Mobile Table II. -cam dimensions of each arthroplasty Kirschner Scorpio NexGen LPS Flex Fixed * Size height Cam Size height Cam Size height Cam XS A S B M C L D XL E F * the size of the post and cam of the NexGen LPS Flex Mobile is identical to that of the NexGen LPS Flex Fixed d Fig. 5 Contact location measurement of the post and cam. The distance to the centre of the contact area was measured from the deepest point of the upper surface of the tibial component (d). again with the tibial component internally rotated by 10 using the Kirschner, Scorpio, NexGen LPS Flex Fixed, and NexGen LPS Flex Mobile components. Contact area, mean and peak contact stress, and contact location were compared for the various components at each angle of flexion. Peak contact stress was also compared between the neutral tibial component and the 10 internally rotated tibial component for each component at each angle of flexion. Statistical significance was calculated with both analysis of variance and a post-hoc Scheffe f test. Values for p < 0.05 were regarded as significant. Results Contact area. This was measured at three different angles, with the various arthroplasties behaving differently. The Kirschner showed the largest contact area, with more than 65 mm 2 at 90 of flexion, an area that was significantly larger than that of both the Scorpio (p = ) and the NexGen (p < ) (Table III). The Kirschner had a decreased contact area at 120 and 150 of flexion, which was significantly smaller than that of the Scorpio (p = (120 ), p < (150 )) and the NexGen (p < (120 ), p = (150 )). The Scorpio had the largest contact area at 150, with up to 60 mm 2, which was significantly larger than the areas for the other arthroplasties (p < ). The NexGen showed minimal changes between 40 to 50 mm 2 for the three different angles of flexion. Mean contact stress. Mean contact stress for the arthroplasties showed an opposing trend to that for contact area. The Kirschner had the lowest contact stress, with less than 8 MPa at 90 of flexion, which was significantly lower than the stresses for the other arthroplasties (p < ) (Table IV). The Kirschner and the Scorpio showed the highest contact stresses at 120, which was significantly higher than the NexGen (p < ). At 150, the Kirschner showed the highest contact stress, higher than both the Scorpio (p < ) and the NexGen (p = ). The Scorpio showed the lowest contact stress at 150, which was significantly lower than the stresses for the other arthroplasties (p < ). The NexGen showed minimal changes with values of between 10 and 12 MPa at the three different angles of flexion. Peak contact stress. Peak contact stresses of between 20 and 35 MPa for the various arthroplasties at the three different angles of flexion were measured. The mean peak stresses were nearly twice as high as the mean contact stresses at the same angle of flexion (Table V). The Kirschner had the lowest contact stress, with less than 25 MPa at 90 of flexion, which was significantly lower than the VOL. 87-B, No. 4, APRIL 2005

4 486 K. NAKAYAMA, S. MATSUDA, H. MIURA, H. HIGAKI, K. OTSUKA, Y. IWAMOTO Table III. Contact area (mm 2 ) of each arthroplasty * ± ± ± 1.4 (60.0 to 72.9) (42.3 to 47.1) (41.3 to 44.5) * ± ± ± 3.2 (34.0 to 34.0) (40.1 to 42.6) (40.3 to 46.7) * ± ± ± 2.4 (40.6 to 43.9) (57.1 to 65.2) (43.9 to 50.3) * significantly different from Scorpio and NexGen (p < 0.05) significantly different from NexGen (p < 0.05) Table VI. The distance from the center of the contact area to the upper surface of the tibia * ± * ± ± 0.5 (5.1 to 5.8) (5.8 to 6.2) (2.4 to 3.4) * ± ± ± 0.2 (6.0 to 6.2) (8.1 to 8.2) (2.6 to 3.1) ± ± ± 0.1 (8.7 to 8.8) (12.1 to 13.4) (8.5 to 8.7) * significantly different from NexGen (p < 0.05) significantly different from Kirschner and NexGen (p < 0.05) Table IV. The mean contact stress (MPa) of each arthroplasty * ± ± ± 0.2 (6.8 to 9.0) (10.9 to 11.3) (11.3 to 11.6) ± ± ± 0.5 (14.5 to 14.9) (14.1 to 15.3) (10.9 to 12.2) * ± ± ± 0.6 (11.4 to 12.5) (7.6 to 9.5) (10.1 to 11.2) * significantly different from Scorpio and NexGen (p < 0.05) significantly different from Kirschner and Scorpio (p < 0.05) Table V. The peak contact stress (MPa) of each arthroplasty ± ± * ± 1.7 (21.8 to 28.3) (24.6 to 28.4) (32.3 to 36.1) ± ± ± 1.0 (33.3 to 34.4) (31.8 to 33.1) (30.0 to 32.4) ± ± * ± 0.7 (27.8 to 29.4) (19.4 to 24.6) (31.8 to 33.2) * significantly different from Kirschner and Scorpio (p < 0.05) significantly different from Scorpio and NexGen (p < 0.05) contact stress for the NexGen (p < ). The Kirschner showed the highest contact stress at 120, which was significantly higher then the stresses for the Scorpio (p = ) and the NexGen (p < ). The Scorpio showed the lowest contact stress at 150, which was significantly lower than that for the other arthroplasties (p < ). Contact location. The centre of the contact area with the Kirschner arthroplasty was 5 to 6 mm in height measured from the deepest part of the upper surface of the tibial component (Table VI). The contact location did not change significantly throughout knee flexion. The location with the Kirschner was significantly higher than for the NexGen at 90 (p < ) and 120 of flexion (p < ). The contact location of the Scorpio was 6.0 mm at 90 of flexion, but moved up to approximately 8 mm at 120 and 150 of flexion. The contact locations at 120 and 150 were significantly higher than those of the Kirschner (p < ) and the NexGen (p < ). The contact location of the Nex- Gen was less than 3 mm and lower than the other arthroplasties at 90 and 120 of flexion, but moved up to almost 6 mm at 150 of flexion. Effect of internal rotation of the tibial components. With internal rotation of the tibial component, the Kirschner, Scorpio, and NexGen LPS Flex Fixed showed an increased peak contact stress compared with neutral rotation at 90 flexion (Table VII). Statistical significance was detected between neutral and internal rotation with the Kirschner (p = ), Scorpio (p < ), and NexGen LPS Flex Fixed (p < ). The NexGen LPS Flex Mobile did not show significantly increased stress with an internally rotated tibial component. At 120 of flexion, internal rotation of the tibial component significantly increased peak contact stresses for both the Kirschner and the NexGen LPS Flex Fixed (p < ) (Table VII). The Scorpio and Nex- Gen LPS Flex Mobile showed relatively small increases in stress with internal rotation. When the arthroplasty was flexed to 150, peak contact stress with the internally rotated tibial component significantly increased with the Kirschner (p = ), Scorpio (p = ), and NexGen LPS Flex Fixed (p < ), but not with the NexGen LPS Flex Mobile arthroplasty (Table VII). Discussion erior-stabilised TKA is one of the most successful procedures in orthopaedic surgery. 2-4 Recently, however, complications with the post-cam mechanism such as fracture or severe wear of the post have been reported. 5-9 Numerous studies have analysed contact stresses at the tibiofemoral and patellofemoral joints in patients with a TKA, in order to achieve better wear characteristics but the post-cam mechanism has not yet been fully investigated. It is known that anteroposterior force is not significantly increased during the gait cycle although biomechanical studies have shown that a higher anteroposterior shear force is applied at the normal tibiofemoral joint during deep knee flexion Dahlkvist et al 12 estimated that between 2.3 and 3.0 times body-weight equivalent of tibiofemoral force is applied tangentially to the tibiofemoral joint when squatting, while Nagura et al 13 reported that deep flexion activities generate 58.3% to 67.8% body-weight equivalent of posterior force. It is clearly necessary therefore, to evaluate the effects of high stress in conditions that stimulate deep flexion. The results of our study show that very high contact stresses, of more than 30 MPa, can be detected at the postcam mechanism of all arthroplasties in deep flexion when THE JOURNAL OF BONE AND JOINT SURGERY

5 CONTACT STRESS AT THE POST-CAM MECHANISM IN POSTERIOR-STABILISED TOTAL KNEE ARTHROPLASTY 487 Table VII. The peak contact stress (MPa) of each arthroplasty with neutral and 10 internally rotated tibial components at 90, 120 and 50 of flexion Component Neutral Internal rotation Neutral Internal rotation Neutral Internal rotation Kirschner 24.0 ± * ± ± * ± ± * ± 3.0 (21.8 to 28.3) (31.6 to 59.7) (33.3 to 34.4) (49.6 to 62.3) (27.8 to 29.4) (31.0 to 38.2) Scorpio 25.9 ± * ± ± ± ± * ± 6.7 (24.6 to 28.4) (36.2 to 45.5) (31.8 to 33.1) (27.9 to 47.4) (19.4 to 24.6) (22.6 to 37.3) NexGen LPS Flex Fixed 34.1 ± * ± ± * ± ± * ± 1.9 (32.3 to 36.1) (44.3 to 52.0) (30.0 to 32.4) (47.0 to 55.6) (31.8 to 33.2) (54.9 to 61.5) NexGen LPS Flex Mobile 34.1 ± ± ± ± ± ± 1.6 (32.3 to 36.1) (27.6 to 39.4) (30.0 to 32.4) (29.8 to 37.1) (31.8 to 33.2) (30.6 to 35.5) * significantly different from neutral (p < 0.05) Fig. 6 The Kirschner arthroplasty with an internally rotated tibial component, at 120 of flexion. The posterolateral edge of the post impinges on the medial edge of the lateral femoral condyle. 500 N of posterior force is applied. These results suggest that severe wear or fracture of the post might occur with the frequent extreme knee bending which is common in daily activities, even with the current designs of TKA, some of which were specifically designed to allow this degree of flexion. Our study also shows significant differences in contact stress for the different arthroplasties at different angles of flexion. The Kirschner TKA increased its contact stress from 90 to 120 of flexion, and then decreased slightly at 150 of flexion. The apex of the cam of the Kirschner is orientated posteriorly, has a relatively flat surface on its distal part, a round shape with small radii posteriorly, and gradually increased radii of curvature posteroproximally (Fig. 1b). These findings suggest that the sharp curve on the posterodistal corner increases contact stress at 120 of flexion. The Scorpio showed higher contact stress than the Kirschner at 90, almost the same stress at 120, and lower stress at 150 of flexion. The cam of the Scorpio has increasing radii of curvature and is relatively flat-shaped posteriorly (Fig. 2b). This explains why this arthroplasty had the largest contact area at 150 of flexion among those tested. The NexGen LPS Flex Fixed did not change contact stress at each angle of flexion. The apex of the cam in the NexGen is orientated posterodistally, and the radii of curvature around the apex do not change significantly (Fig. 3b). This allows the arthroplasty to have a constant contact stress throughout flexion from 90 to 150. The contact location between post and cam is determined by many factors, such as the shape of the cam, the position of attachment of the cam to the femoral component, and the curvature of the posterior femoral condyle. The contact location with the NexGen LPS Flex remained lower than the other arthroplasties at 90 and 120 of flexion. This feature is beneficial for avoiding excessive stress at the bone-implant interface and to prevent fracture of the post. The Kirschner did not change its contact location significantly during flexion, which is theoretically advantageous in avoiding excessive wear. However, we need to investigate the component under dynamic conditions in order to confirm this theory. The effect of tibial rotation is an important issue for post and cam contact mechanisms. The tibia is expected to internally rotate with knee flexion both in the normal knee and after TKA. 23 In our study, very high contact stress was observed at the post-cam mechanism of all fixed-bearing arthroplasties when the tibial component was rotated. Stress increased more than twofold compared with neutral rotation. Significant differences still existed, however, between the different arthroplasties. It appeared that contact configuration with a rotated tibia depended upon two factors. First, the of the post relative to the cam and, secondly the shape of the post. The Kirschner arthroplasty showed a significantly increased peak contact stress with a rotated tibia at 90 and 120 of flexion, but stress increased by only a relatively small amount at 150. The Kirschner post-cam has a curved configuration, which is supposed to accommodate rotation of the tibia. However, it also has a relatively wide post, causing impingement of the edge of the post against the inner surface of the femoral condyle (Fig. 6). By VOL. 87-B, No. 4, APRIL 2005

6 488 K. NAKAYAMA, S. MATSUDA, H. MIURA, H. HIGAKI, K. OTSUKA, Y. IWAMOTO 150 of flexion, the post does not impinge on the femoral component because of the shallow intercondylar space. The Scorpio increased its contact stress with a rotated tibia, but by a smaller amount than for the other fixed designs. It, too, has a curved post-cam configuration, but the post is sufficiently narrow to avoid impingement. The NexGen Flex Fixed has a flat post-cam design and demonstrated the greatest increase in stress because of edge loading. These results suggest that a flat-shaped post may cause severe wear at the post-cam mechanism during flexion. The role of a mobile bearing on the post-cam contact mechanism is another important issue. Mobile-bearing TKAs were developed in order to produce near-normal kinematics and to have a large contact area. 24,25 Biomechanical studies have shown its theoretical advantages in tibiofemoral contact stress and knee kinematics. 15,26 However, no study has investigated the effect of a mobile bearing on the post-cam contact mechanism. The results of our study show that a mobile bearing prevents increased contact stress during tibial rotation and suggest that rotational mobility is important for avoiding severe wear of the post for a knee arthroplasty which has a flat-shaped post-cam design. Our study has some limitations. The tests were performed on the assumption that all stress was applied at the post. -operatively, however, the femoral component does not always make proper contact with the tibial component (lift-off). 27 It is, therefore, difficult to determine the most appropriate situation in which to measure contact stresses at the post-cam mechanism. In our study, a relatively low force of 500 N was chosen and any force applied to the tibiofemoral articular surface was ignored. Regardless of these shortcomings, we believe that our study has revealed important information about the post-cam contact mechanism. It appears that very high contact stress is inevitable at the post-cam mechanism with the current designs of posteriorstabilised TKA components, especially during deep knee flexion. This would suggest that deep bending of the knee should not be recommended for patients after a posterior-stabilised TKA. The shape and apical orientation of the cam can affect contact stress significantly. We consider that the post-cam design should be modified in order to provide a larger contact area whilst avoiding impingement and edge loading. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References 1. Insall JN, Hood RW, Flawn LB, Sullivan DJ. The total condylar knee prosthesis in gonarthrosis: a five to nine-year follow-up of the first one hundred consecutive replacements. J Bone Joint Surg [Am] 1983;65-A: Colizza WA, Insall JN, Scuderi GR. The posterior stabilized total knee prosthesis: assessment of polyethylene damage and osteolysis after a ten-year-minimum followup. J Bone Joint Surg [Am] 1995;77-A: Schai PA, Thornhill TS, Scott RD. Total knee arthroplasty with the PFC system: results at a minimum of ten years and survivorship analysis. J Bone Joint Surg [Br] 1998;80-B: Rodriguez JA, Bhende H, Ranawat CS. Total condylar knee replacement: a 20- year followup study. Clin Orthop 2001;388: Mikulak SA, Mahoney OM, dela Rosa MA, Schmalzried TP. Loosening and osteolysis with the press-fit condylar posterior-cruciate-substituting total knee replacement. J Bone Joint Surg [Am] 2001;83-A: Mauerham DR. Fracture of the polyethylene tibial post in a posterior cruciate-substituting total knee arthroplasty mimicking patellar clunk syndrome: a report of 5 cases. J Arthroplasty 2003;18: Puloski SK, McCalden RW, MacDonald SJ, Rorabeck CH, Bourne RB. Tibial post wear in posterior stabilized total knee arthroplasty: an unrecognized source of polyethylene debris. J Bone Joint Surg [Am] 2001;83-A: Mestha P, Shenava Y, D Arcy JC. Fracture of the polyethylene tibial post in posterior stabilized (Insall Burstein II) total knee arthroplasty. J Arthroplasty 2000;15: Callaghan JJ, O Rourke MR, Goetz DD, et al. Tibial post impingement in posterior-stabilized total knee arthroplasty. Clin Orthop 2002;404: Sultan PG, Most E, Schule S, Li G, Rubash HE. Optimizing flexion after total knee arthroplasty: advances in prosthetic design. Clin Orthop 2003;416: Matsuda S, Whiteside LA, White SE, McCarthy DS. Knee kinematics of posterior cruciate ligament sacrificed total knee arthroplasty. Clin Orthop 1997;341: Dahlkvist NJ, Mayo P, Seedhom BB. Forces during squatting and rising from a deep squat. Eng Med 1982;11: Nagura T, Dyrby CO, Alexander EJ, Andriacchi TP. Mechanical loads at the knee joint during deep flexion. J Orthop Res 2002;20: Wilk KE, Escamilla RF, Fleisig GS, et al. A comparison of tibiofemoral joint forces and electromyographic activity during open and closed kinetic chain exercises. Am J Sports Med 1996;24: Li G, Zayontz S, Most E, et al. In situ forces of the anterior and posterior cruciate ligaments in high knee flexion, and in vitro investigation. J Orthop Res 2004;22: Li G, Zayontz S, DeFrate LE, et al. Kinematics of the knee at high flexion angles: an in vitro investigation. J Orthop Res 2004;22: Li G, Most E, Otterberg E, et al. Biomechanics of posterior-substituting total knee arthroplasty: an in vitro study. Clin Orthop 2002;404: Matsuda S, White SE, Williams VG 2nd, McCarthy DS, Whiteside LA. Contact stress analysis in meniscal bearing total knee arthroplasty. J Arthroplasty 1998;13: Matsuda S, Ishinishin T, White SE, Whiteside LA. Patellofemoral joint after total knee arthroplasty: effect on contact area and contact stress. J Arthroplasty 1997;12: Matsuda S, Whiteside LA, White SE. The effect of varus tilt on contact stresses in total knee arthroplasty: a biomechanical study. Orthopaedics 1999;22: Lee TQ, Gerken AP, Glaser FE, Kim WC, Anzel SH. Patellofemoral joint kinematics and contact pressures in total knee arthroplasty. Clin Orthop 1997;340: Stukenborg-Colsman C, Ostermeier S, Hurschler C, Wirth CJ. Tibiofemoral contact stress after total knee arthroplasty: comparison of fixed and mobile-bearing inlay designs. Acta Orthop Scand 2002;73: Most E, Li G, Schule S, et al. The kinematics of fixed- and mobile-bearing total knee arthroplasty. Clin Orthop 2003;416: Goodfellow JW, O Connor J. Clinical results of the Oxford knee: surface arthroplasty of the tibiofemoral joint with a meniscal bearing prosthesis. Clin Orthop 1986; 205: Buechel FF, Pappas MJ. The New Jersey low-contact-stress knee replacement system: biomechanical rationale and review of the first 123 cemented cases. Arch Orthop Trauma Surg 1986;105: Matsuda S, Whiteside LA, White SE, McCarthy DS. Knee stability in meniscal bearing total knee arthroplasty. J Arthroplasty 1999;14: Dennis DA, Komistek RD, Walker SA, Cheal EJ, Stiehl JB. Femoral condylar liftoff in vivo in total knee arthroplasty. J Bone Joint Surg [Br] 2001;83-B:33-9. THE JOURNAL OF BONE AND JOINT SURGERY