The effect of femoral component rotation on the kinematics ... - umexpert

0 downloads 0 Views 586KB Size Report
Apr 12, 2011 - external rotation of 5° caused tibial abduction and then adduction of 5° ..... medial translation between 08 and 508 knee flexion, and vice versa ...
Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487 DOI 10.1007/s00167-011-1499-8

KNEE

The effect of femoral component rotation on the kinematics of the tibiofemoral and patellofemoral joints after total knee arthroplasty A. M. Merican • K. M. Ghosh • F. Iranpour D. J. Deehan • Andrew A. Amis



Received: 23 September 2010 / Accepted: 21 March 2011 / Published online: 12 April 2011 Ó Springer-Verlag 2011

Abstract Purpose Complications after total knee arthroplasty (TKA) often involve the patellofemoral joint, and problems with patellar maltracking or lateral instability have sometimes been addressed by external rotation of the femoral component. This work sought to measure the changes of knee kinematics caused by TKA and then to optimise the restoration of both the patellofemoral and tibiofemoral joint kinematics, by variation of femoral component internal–external rotation. Methods The kinematics of the patella and tibia were measured in eight cadaveric knees during active extension motion. This was repeated with the knee intact, with a Genesis II TKA in the standard position (3° of external rotation) and with the femoral component at ±5° rotation from there. Results Both patellar and tibial motions were significantly different from normal with the standard TKA rotation, with 3° tibial abduction at 90° flexion and reversal of A. M. Merican  F. Iranpour Department of Orthopaedic Surgery, Imperial College London, South Kensington Campus, London SW7 2AZ, UK K. M. Ghosh  A. A. Amis (&) Department of Mechanical Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK e-mail: [email protected] A. M. Merican Department of Orthopaedic Surgery, University of Malaya Medical Centre, Kuala Lumpur, Malaysia K. M. Ghosh  D. J. Deehan Department of Orthopaedic Surgery, Newcastle University Hospital, Newcastle-Upon-Tyne, UK

the screw-home from 5° external rotation to 6° internal rotation. The patella was shifted medially 6 mm in flexion and tilted 7° more laterally near extension. Femoral rotation to address one abnormality caused increased abnormality in other degrees of freedom. Internal and then external rotation of 5° caused tibial abduction and then adduction of 5° at 90° flexion. These femoral rotations also caused increased patellar lateral tilt of 4° with femoral external rotation and decreased tilt by 4° with internal rotation. Thus, correction of tibial abduction in flexion, by external rotation of the femoral component, worsened the patellar lateral tilt near extension. Conclusions It was concluded that femoral rotation alone could not restore all aspects of both patellar and tibial kinematics to normal with this specific implant. The clinical relevance of this is that it appears to be inadvisable to reposition the femoral component, in an attempt to improve patellar tracking, if that repositioning may then cause abnormal tibiofemoral kinematics. Further, the pattern of patellar tracking, with the type of TKA used in this study, could not be adjusted to normal by femoral component rotation. Keywords TKA  Femoral component rotation  Knee kinematics  Patellar tracking

Introduction Patellofemoral complications are a major cause of poor function in the prosthetic knee [6, 7, 14, 16]. There is good experimental and clinical evidence that poor femoral component rotational alignment can adversely affect patellar tracking and kinematics [3, 12, 20, 23, 26]. While some studies found that femoral component rotation affects

123

1480

lateral release rates [1, 27], a prospective randomised trial did not show a difference [30], and so lateral release may not be appropriate in this situation. The Genesis II implant has an external rotation built into the posterior femoral condyles, and the consequent lateralised trochlear groove near extension is intended to facilitate good patellar tracking. Kaper et al. [15] found that their lateral release rate dropped from 14 to 5% when the Genesis II was used instead of the Genesis I, but kinematic analysis of the Genesis II has demonstrated significant lateral tilting [6]. The effect of differing femoral component rotations on both tibiofemoral and patellofemoral kinematics was, however, not investigated. The prevalence of less than normal function after TKA shows that there is a need for more data to describe the effects of total knee arthroplasty and particularly the effects of differing choices for the rotational position of the femoral component. The clinical situation is confused by the variable relationship between the tibiofemoral and patellofemoral articulations [10] and by the variability of defining the transepicondylar axis [14], which is often used as a datum for femoral component rotation. There will be interactions between the methods used to tense and balance the collateral ligaments, the resurfacing of the patella and the releases of the patellofemoral retinacula, and so it is not surprising that this area remains controversial. The aim of this study was to investigate how the kinematics of both the tibiofemoral and patellofemoral joints are altered by implantation of the Genesis II TKA and also to study the effects of altering the internal–external rotational position of the femoral component. It was hypothesised that external rotation of the femoral component would cause the tibia to move into external rotation with the knee extended and into varus when flexed, and vice versa. Similarly, it was hypothesised that the patella would be translated and tilted laterally by femoral component external rotation, and vice versa, near knee extension. This led to the further hypothesis that, once these kinematic changes were found, the optimal orientation of the femoral component in internal–external rotation could be identified, respecting both articulations.

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

changes sufficient to have an effect on the alignment or kinematics. The study was approved by the Riverside Research Ethics Committee. The knees were stored at -20°C and thawed a day prior to experimentation. The skin and subcutaneous tissue were removed. The deep fascia, retinacula and iliotibial band (ITB) were preserved. The femur and tibia were cut approximately 20 and 15 cm above and below the knee, respectively. The head of the fibula was transfixed to the tibia by two bone screws in its anatomical position. An intramedullary sleeve and a rod were cemented into the femur and tibia, respectively, using polymethylmethacrylate. They were accurately aligned to the anatomical axes by use of rubber spacers and an outrigger alignment tool. This custom-made instrument engaged the intramedullary rod where it emerged from the centre of the cut end of the bone, and its other end was placed so that it overlaid the anatomical axis position on the femur or tibia at the knee joint line. The quadriceps was separated into six components: rectus femoris (RF), vastus intermedius (VI), vastus lateralis longus (VLL), vastus lateralis obliquus (VLO), vastus medialis longus (VML) and vastus medialis obliquus (VMO). The knee was mounted in a rig by sliding the femoral sleeve onto a fixed rod (Fig. 1). The femur was fixed with the most-posterior parts of the femoral condyles horizontal. The components of the quadriceps and the iliotibial band (ITB) were each loaded with hanging weights using cables and pulleys with a total of 175 N [9] and 30 N [5, 21], respectively, according to the physiological cross-sectional areas and directions of the muscles relative to the femoral axis [8]. The direction of pull of the ITB was 08 lateral and 68 posterior to the femoral axis [5].

Methods Specimen preparation Eight fresh-frozen cadaveric knees aged 63 ± 16 years (mean ± SD) were used in this study. They were obtained from the International Institute for the Advancement of Medicine (Jessup, PA, USA), who undertook screening and consenting for their use for research. The knees had no signs of previous surgery, varus-valgus deformity, ligament laxity or injury and did not have articular degenerative

123

Fig. 1 A right knee in the kinematics test rig, with the distal end of the tibia in extended posture at the bottom, left corner of the picture. The tibia is free to flex-extend in the vertical plane, while the femur (not visible beneath the quadriceps) is fixed to the rig. Each of the components of the quadriceps is loaded by strings passing over one of the pulleys at the top, right of the picture, proximal to the specimen. Each muscle head is tensed by hanging weights, which are not visible. Optical trackers are secured to the femur (F), patella (P) and tibia (T)

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

Tracking A Polaris optical system (Northern Digital Incorporated, Waterloo, Canada) was used with active optical trackers (Fig. 1). The Polaris system had an overall volume root mean square (RMS) distance error of 0.35 mm for a single marker [29]. The Traxtal tracker used for the patella had a mean accuracy of 0.038 and 0.04 mm [13, 21]. The knee was moved into two cycles of flexion–extension, against the extending moment of the quadriceps tension, using a rod held transversely against the anterior surface of the tibial intramedullary rod. Both of these rods were cylindrical, and so the tibia was unconstrained and free to rotate. Data collected during knee extension were saved for analysis. The femoral coordinate system was aligned to the anatomical axis and the most-posterior points of the femoral condyles. The origin of the patellar coordinate system was 10 mm deep to the anterior surface, overlying the proximal–distal centre of the median ridge, aligned to the medial, lateral and distal points in the transverse and coronal planes. For patellar lateral translation (or shift), the position in extension in the native knee was designated as 0 mm; then, measurements were made perpendicular to the anatomical axis and parallel to the plane of the posterior condyles. The tibial coordinate system was based on the anatomical axis and the most medial and lateral points of the tibial plateau [19]. The anatomical landmarks were digitised using a stylus to relate them to the trackers. The raw kinematic data were processed with Visual3D (C-Motion Inc., Maryland, USA). The motion of the tibia in relation to the femur was calculated according to the joint coordinate system for the knee [11]. Lateral

Fig. 2 Definitions of the individual degrees-of-freedom of patellar motion

1481

patellar tilt was defined as a rotation about the longitudinal axis of the patella with positive values indicating that the lateral patella approached the femur. Positive patellar lateral rotation means that the distal patella moved laterally relative to its centre (Fig. 2). Surgical technique A patellar-splitting approach was used, preserving the retinacula, hence avoiding inconsistencies in a medial parapatellar closure which might affect tracking of the patella [22]. After the natural knee kinematics was measured, the optical trackers were removed; their mounts allowed them to return to their original position after TKA. The TKA was a posterior cruciate ligament retaining Genesis II (Smith and Nephew, Memphis TN). The tibial cut was made perpendicular to the anatomical axis in the coronal plane and with a 38 posterior slope using an extramedullary jig that was located distally against the protruding end of the tibial intramedullary rod. Bone cuts were made as a measured resection technique, and no ligament release was performed; the posterior condylar line was used as the datum for femoral component rotation and was set accurately horizontal in the kinematics rig using a spirit level pressed against the posterior aspect of the knee. The femoral implant had a thinner posteromedial condyle compared to the posterolateral condyle. This conferred an in-built external rotation in flexion. Hence when the femoral component was placed in neutral rotation with respect to the posterior condyles, the flexion space was balanced mediolaterally with a neutral coronal tibial cut without having to externally rotate the femoral component. Extra bone was then removed from the femur to allow 58 of internal and external rotation from this datum to be simulated. A conversion module (from the revision instrumentation) was secured to an intramedullary rod. The femoral component would later be cemented to this module. This rod was passed through the femoral intramedullary sleeve and was accessible at the proximal end of the rig. Thus, the femoral component could be rotated within the closed knee. A protractor and pointer were attached to the proximal end of the rod and mounting. A spirit level was used to position the femoral component in neutral rotation (Fig. 3). The joint line position was controlled by measuring the level of the distal articular surface from a femoral fiducial marker. The tibial component was cemented with its rotation based on the junction of the medial and middle thirds of the tuberosity, avoiding internal rotation. The thickness of the patella was measured using a digital calliper. It was then cut using a jig, and the same thickness was restored with the patellar prosthesis within ±0.5 mm. The patellar component was centred at the previously measured midpoint of the median ridge, which had been

123

1482

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

Fig. 3 The femoral conversion module attached to the intramedullary rod. A spirit level is used to obtain neutral rotation with respect to the femoral posterior condylar axis. The femoral component is cemented over the module and the knee closed. This conversion module, with the femoral component mounted on it, could then be rotated into 58 internal–external rotation from neutral rotation within the closed knee, from the proximal end of the intramedullary rod

marked by a small drill hole prior to the resection. The patellar split was closed using two cannulated cancellous bone screws in predrilled holes. The cut in the quadriceps tendon was approximated without tension using a suture. The trackers were replaced on the bones. The kinematic measurements were repeated. The intramedullary rod attached to the femoral component was unlocked and rotated the femoral component 58 internally from the datum position without reopening the joint. The kinematic measurements were repeated. This was repeated again after the femoral component was rotated externally by 58 from the datum position. Statistical analysis The kinematic data were analysed at every 58 of knee flexion using repeated-measures two-way analyses of variance (ANOVA), examining differences before and after the TKA and between neutral rotation and the two malrotations. Bonferroni post hoc tests were used to determine in which testing conditions and knee flexion angles, there was a statistically significant difference compared to the native knee. Differences were taken to be significant for P \ 0.05.

123

Fig. 4 Tibial abduction: a For the native knee and after knee replacement. b The effect of 58 internal (TKRIR5) and external rotation (TKRER5) of the femoral component compared to the knee replacement in neutral rotation (TKR) (Mean ± SD, n = 8)

Results In extension, the tibial abduction angle before and after TKA was 58 ± 28 (mean ± SD) in relation to the femoral anatomical axis. The replaced knee was 38 ± 28 more abducted than the native knee at 908 flexion (P \ 0.05) (Fig. 4a). Internal and external rotation of the femoral component caused tibial abduction and adduction, respectively, from 358 flexion onwards (Fig. 4b). At 908 flexion, 5° internal rotation of the femoral component caused 58 ± 38 tibial abduction, and 5° external rotation caused 58 ± 38 tibial adduction.

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

1483

Fig. 5 Tibial external rotation: a Tibial external rotation relative to the femur for the native knee and after knee replacement. b The effect of internal (TKRIR5) and external rotation (TKRER5) of the femoral component by 58 compared to the knee replacement in neutral rotation (TKR). (Mean ± SD, n = 8)

Fig. 6 Patellar lateral translation: a Patellar lateral translation relative to the femur for the native knee and after knee replacement. b The effect of internal (TKRIR5) and external rotation (TKRER5) of the femoral component by 58 compared to the knee replacement in neutral rotation (TKR). (Mean ± SD, n = 8)

For tibial internal–external rotation, all of the intact knees displayed an external rotation (‘screw-home’) of 58 ± 38 in the final 188 ± 58 knee extension (Fig. 5a). This pattern was reversed after TKA, with an internal rotation of 68 ± 38 in the last 208 ± 68 extension. In extension, the tibia was 5° ± 28 internally rotated after TKA, compared with the natural knee (P \ 0.001). Conversely, the tibia was significantly more externally rotated after TKA between 158 and 458 flexion; by 68 ± 48 at 258 flexion. Malrotation of the femoral component caused significant changes of tibial rotation between 08 and 658 of flexion (Fig. 5b). Internal rotation of the femoral

component caused tibial internal rotation, and vice versa, reaching 48 ± 28 at 308 of flexion. The patellar motion for the natural knee had an initial medial translation of 2 ± 1 mm from 08 to 158 flexion (P \ 0.05) followed by lateral translation of 5 ± 2 mm (P \ 0.001) by 908 flexion (Fig. 6a). After TKA, the patella was 4 ± 3 mm (P \ 0.01) more medial than in the native knee at 08 flexion. A significant difference was not shown between 58 and 608. Beyond 658 flexion, the patella was more medial post-TKA compared to the intact knee, reaching 6 ± 3 mm at 958 flexion (Fig. 6a). Internal rotation of the femoral component caused significant patellar

123

1484

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

medial translation between 08 and 508 knee flexion, and vice versa (Fig. 6b), reaching 3 ± 1 mm medially at 308 knee flexion, and 4 ± 2 mm laterally at 208 knee flexion. In the native knee, patellar tilting followed a pattern similar to that for lateral translation (Fig. 7a). The lateral tilt of 68 ± 38 in extension increased to 128 ± 28 at 90° knee flexion (P \ 0.01). After TKA, there was a significant increase of patellar lateral tilt between 108 and 458 knee flexion (Fig. 7a), reaching 78 ± 58 more lateral tilt at 208 flexion. Internal rotation of the femoral component decreased the patellar lateral tilt between 58 and 758 of

flexion, reaching 48 ± 38 at 308 knee flexion. External rotation of the femoral component increased patellar lateral tilt between 08 and 858, reaching 48 ± 28 at 408 knee flexion. Patellar lateral rotation decreased with knee flexion before and after TKA, but it was more laterally rotated after TKA beyond 508 flexion, reaching 48 ± 38 at 758 knee flexion (Fig. 8a). Malrotation of the femoral component did not change patellar rotation when the knee was extended, but caused significant changes above 358 knee flexion (Fig. 8b). At 908 knee flexion, external rotation of

Fig. 7 Patellar lateral tilt: a Patellar lateral tilt relative to the femur for the native knee and after knee replacement. b The effect of internal (TKRIR5) and external rotation (TKRER5) of the femoral component by 58 compared to the knee replacement in neutral rotation (TKR). (Mean ± SD, n = 8)

Fig. 8 Patellar lateral rotation: a Patellar lateral rotation relative to the femur for the native knee and after knee replacement. b The effect of internal (TKRIR5) and external rotation (TKRER5) of the femoral component by 58 compared to the knee replacement in neutral rotation (TKR). (Mean ± SD, n = 8)

123

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

the femoral component caused 48 ± 28 medial rotation of the patella, and vice versa. In the intact knee, patellar flexion increased linearly from 48 ± 58 at 08 knee flexion to 718 ± 68 at 908 knee flexion. After TKA, the patella tended to be more flexed than in the intact knee near extension, becoming less flexed than in the intact knee with increasing knee flexion, the difference reaching 48 ± 38 at 808 flexion (P \ 0.05). No significant effect on patellar flexion was found from femoral component rotation.

Discussion The most important finding of this study was that the kinematics of the prosthetic and native knees differed significantly, for both the tibiofemoral and patellofemoral joints, and that femoral component rotation could not be used to adjust them both simultaneously towards normal. The kinematic effects of malrotation of the femoral component conformed to the original hypotheses: with external rotation of the femoral component, the tibia was carried into external rotation near extension, and adduction in flexion. Similarly, near knee extension, the patella was translated and tilted laterally by femoral component external rotation. However, contrary to our second hypothesis, the conflicting effects caused by femoral component rotation, between the different degrees-of-freedom of the patellar and tibial kinematics, meant that it was difficult to say which position would be best for simultaneously restoring the kinematics of both the patellofemoral and tibiofemoral joints as closely as possible to normal. It would have been very useful if this experiment had been able to identify a set-up that gave the closest approach to normal kinematics in both joints, after which only small adjustments such as soft tissue releases might still be needed. Unfortunately, that was not the case with the prosthesis design used in this study. With growing efforts to optimise the set-up of TKA, we foresee that data similar to that derived in this study could become the basis for adjusting the settings of surgical alignment instruments, or the guidance provided in surgical navigation systems. For tibiofemoral kinematics, the important differences between the native knee and the prosthetic knee were as follows: the loss of the normal screw-home mechanism when the knee was extended; the tibia was more externally rotated than the native knee between 158 and 458 flexion; and after 658 flexion, the tibia was more abducted in the prosthetic knee. Internal rotation of the prosthetic knees in late flexion suggests that posterior femoral rollback on the lateral side was preserved. Chew et al. [6] studied the tibiofemoral kinematics before and after implantation of the Genesis II prosthesis. Despite different muscle tensions, their results were similar to this study: the native and

1485

prosthetic knees had decreasing tibial abduction with knee flexion, and there was loss of the screw-home mechanism after TKA. The prosthetic knee had less external rotation than in this study, which loaded the ITB, increasing tibial external rotation [21]. The pattern of patellar medial–lateral translation was altered by TKA, with the implant taking the patella more medial in deep flexion. The patellar lateral tilt was also increased by the prosthesis between 108 and 458 knee flexion. This corresponded to the knee flexion angles in which the tibia was more externally rotated. Chew and colleagues [6] also found 7° lateral tilt with the Genesis II, at 608 knee flexion. We postulate that the lateral patellar tilt after TKA was related to the alteration of the tibiofemoral kinematics. The graphs of lateral tilt and of tibial external rotation versus knee flexion had similar patterns, in the native knee and the prosthetic knee, with the patella tilting to follow the tibial rotation, as in previous studies [17, 24, 28]. Generally, the changes in kinematics after malrotation of the femoral component were predictable and as we had hypothesised. With external rotation of the femoral component, the tibia rotated into adduction in flexion and external rotation in extension, because of the articular constraint. At 908 flexion, the change in abduction and adduction corresponded to the amount of femoral malrotation. The change in tibial internal–external rotation was expected to be maximal at 08 flexion, but this was observed around 308 flexion, probably reflecting the slackening of posterior tissues in early flexion. This effect diminished with increasing flexion, as expected. Similarly, with the patella, changes in lateral rotation occurred more in flexion and changes of patellar tilt near extension. External rotation of the femoral component caused the trochlear groove to translate laterally and rotate externally, and these changes were mirrored by the patella, which was carried with it, as in previous work [17, 24, 28]. There are strengths and limitations of this in vitro study which influence its clinical relevance. The quadriceps components and ITB were loaded in physiological directions. Antagonistic hamstrings loading was omitted because the extension part of the cycle was used, and there is little data on the relative tension during knee motion. Although hamstrings tension affects the kinematics [18], we felt that it was not practical to add an investigation of their effects to an already long experimental protocol, because of our concern about deterioration of the specimens. Previous works on knee extension have shown the importance of the oblique components of the quadriceps [4] and using multiplanar loading directions on patellar mechanics [25]. Although the individual heads of the quadriceps were tensioned in a physiological ratio, their overall tension was limited by tearing the muscle fibres in these elderly specimens. A laboratory experiment utilising

123

1486

normal knees may achieve better accuracy than in real-life surgery because of the normal bony landmarks, absence of deformity and the possibility of making measurements and referencing cuts not possible in clinical surgery. Correct rotational alignment of the femoral prosthesis was ensured by fine adjustment to achieve rotation neutral to the posterior condylar axis. The use of cadaveric knees for sequences of implant positions allowed repeated-measures statistical analysis, which eliminates variability such as soft tissue or other pathological or functional differences between patients that limit the power of clinical studies of implant positioning. In addition to these strengths and weaknesses, it is known that knee kinematics when loadbearing in vivo is altered by implant design, patient activity and individual surgeon technique [2]. This study has concentrated on tibiofemoral alignment and patellar kinematics rather than aspects such as AP translation, which also will be affected by TKA with ACL resection. This study measured both tibiofemoral and patellofemoral kinematics simultaneously, in order to understand their interactions. It had been hypothesised that this work would indicate how best to position the femoral component in internal–external rotation in order to optimise the restoration of normal kinematics of both joints simultaneously. However, comparison of the graphs suggests that it will not be possible to do this by this manoeuvre alone: there are conflicting findings for both joints. For the tibia, Fig. 4 suggests that tibial abduction in flexion would be optimised if the femoral component were to be externally rotated 5° from the neutral rotation position. Conversely, Fig. 5 shows that there was excess tibial external rotation from 20° to 50° flexion, which could be reduced by internal rotation of the femoral component. There is a similar conflict for the patella: Fig. 6 shows that the patella tracked medial to the natural path post-arthroplasty, which could be partly corrected by external rotation of the femoral component, while Fig. 7 shows that there was also lateral tilt, which would be partly corrected by internal rotation of the femoral component. Thus, the kinematics of both articulations could probably not be returned to normal by adjusting the rotation of the femoral component. This conflict is reflected by the range of investigations and recommendations in the literature, from external rotation [26] to neutral to the transepicondylar axis [23]. Moreover, the relationships between the patellar groove and rotational reference axes of the femur vary considerably between individuals [10]. The clinical relevance of this study relates to the demonstration of how the patellofemoral and tibiofemoral joint kinematics are inter-related via the femoral component after TKA. Thus, an alteration of one factor alters the kinematics of the other joint. In particular, the femoral component is often rotated away from being parallel to the

123

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487

trans-epicondylar axis in an attempt to improve the tracking characteristics of the patella, but this work has shown how there were contrary effects in other degrees-of-freedom. The geometry of the specific TKA used in this study did not allow us to optimise all aspects of knee kinematics simultaneously.

Conclusion This study found that, with the prosthesis studied, we could not replicate natural kinematics by adjusting femoral component rotation. Patellar motion was coupled to tibial kinematics, so external rotation of the femoral implant by an additional 5° on top of the built-in external rotation, in an attempt to correct the tibial abduction in flexion and the patellar medial translation, would have increased the abnormal tibial external rotation and patellar lateral tilt. It appears to be unlikely that, with this specific implant design, adjustment of the femoral component rotation alone could restore kinematics to normal post-arthroplasty. Acknowledgments K. Milton Ghosh and the running costs of this work were supported by Smith and Nephew (Reconstructive) Ltd., UK. Azhar M Merican was supported by the University of Malaya Medical Centre, Kuala Lumpur and the Arthritis Research (UK) charity. Farhad Iranpour was supported by the Furlong Charitable Research Foundation. We thank W. Scott Selbie PhD, C-Motion Inc., for his software support. None of the authors have any financial or commercial links that may be perceived to have biased this article.

References 1. Akagi M, Matsusue Y, Mata T, Asada Y, Horiguchi M, Iida H, Nakamura T (1999) Effect of rotational alignment on patellar tracking in total knee arthroplasty. Clin Orthop Relat Res 366:155–163 2. Banks SA, Harman MK, Bellemans J, Hodge WA (2003) Making sense of knee kinematics: news you can use. J Bone Jt Surg (Am) 85:64–72 3. Berger RA, Crossett LS, Jacobs JJ, Rubash HE (1998) Malrotation causing patellofemoral complications after total knee arthroplasty. Clin Orthop Relat Res 356:144–153 4. Bose K, Kanagasuntheram R, Osman MBH (1980) Vastus medialis oblique—an anatomic and physiologic study. Orthopedics 3:880–883 5. Bull AM, Andersen HN, Basso O, Targett J, Amis AA (1999) Incidence and mechanism of the pivot shift. An in vitro study. Clin Orthop Relat Res 363:219–231 6. Chew JTH, Stewart NJ, Hanssen AD, Luo ZP, Rand JA, An KN (1997) Differences in patellar tracking and knee kinematics among three different total knee designs. Clin Orthop Relat Res 345:87–98 7. Dalury DF, Dennis DA (2003) Extensor mechanism problems following total knee replacement. J Knee Surg 16:118–122 8. Farahmand F, Senavongse W, Amis A (1998) Quantitative study of the quadriceps muscles and trochlear groove geometry related to instability of the patellofemoral joint. J Orthop Res 16:136–143

Knee Surg Sports Traumatol Arthrosc (2011) 19:1479–1487 9. Farahmand F, Tahmasbi MN, Amis AA (1998) Lateral forcedisplacement behaviour of the human patella and its variation with knee flexion–a biomechanical study in vitro. J Biomech 31:1147–1152 10. Feinstein WK, Noble PC, Kamaric E, Tullos HS (1996) Anatomic alignment of the patellar groove. Clin Orthop Relat Res 331:64– 73 11. Grood ES, Suntay WJ (1983) A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. J Biomech Eng 105:136–144 12. Heegaard JH, Leyvraz PF, Hovey CB (2001) A computer model to simulate patellar biomechanics following total knee replacement: the effects of femoral component alignment. Clin Biomech 16:415–423 13. Iranpour F, Merican AM, Baena FR, Cobb JP, Amis AA (2010) Patellofemoral joint kinematics: the circular path of the patella around the trochlear axis. J Orthop Res 28:589–594 14. Jerosch J, Peuker E, Philipps B, Filler T (2002) Inter-individual reproducibility in perioperative rotational alignment of femoral components in knee prosthetic surgery using the transepicondylar axis. Knee Surg Sports Traumatol Arthrosc 10:194–197 15. Kaper BP, Woolfrey M, Bourne RB (2000) The effect of built-in external femoral rotation on patellofemoral tracking in the Genesis II total knee arthroplasty. J Arthroplast 15:964–969 16. Kelly MA (2001) Patellofemoral complications following total knee arthroplasty. Instr Course Lect 50:403–407 17. Kessler O, Colwell CW, D’Lima DD (2008) The effect of femoral component rotation on patellar biomechanics. J Biomech 41:3332–3339 18. Kwak SD, Ahmad CS, Gardner TR, Grelsamer RP, Henry JH, Blankevoort L, Ateshian G, Mow VC (2000) Hamstrings and iliotibial band forces affect knee kinematics and contact pattern. J Orthop Res 18:101–108 19. Lie DT, Bull AM, Amis AA (2007) Persistence of the mini pivot shift after anatomically placed anterior cruciate ligament reconstruction. Clin Orthop Relat Res 457:203–209

1487 20. Malo M, Vince KG (2003) The unstable patella after total knee arthroplasty: etiology, prevention, and management. J Am Acad Orthop Surg 11:364–371 21. Merican AM, Amis AA (2009) Iliotibial band tension affects patellofemoral and tibiofemoral kinematics. J Biomech 42:1539– 1546 22. Merican AM, Ghosh KM, Deehan DJ, Amis AA (2009) The transpatellar approach for the knee in the laboratory. J Orthop Res 27:330–334 23. Miller MC, Berger RA, Petrella AJ, Karmas A, Rubash HE (2001) The Ranawat award paper—optimizing femoral component rotation in total knee arthroplasty. Clin Orthop Relat Res 392:38–45 24. Nagamine R, Whiteside LA, Otani T, White SE, McCarthy DS (1996) Effect of medial displacement of the tibial tubercle on patellar position after rotational malposition of the femoral component in total knee arthroplasty. J Arthroplast 11:104–110 25. Powers CM, Lilley JC, Lee TQ (1998) The effects of axial and multi-plane loading of the extensor mechanism on the patellofemoral joint. Clin Biomech 13:616–624 26. Rhoads DD, Noble PC, Reuben JD, Mahoney OM, Tullos HS (1993) The effect of femoral component position on patellar tracking after total knee arthroplasty. Clin Orthop Relat Res 260:43–51 27. Sodha S, Kim J, McGuire KJ, Lonner JH, Lotke PA (2004) Lateral retinacular release as a function of femoral component rotation in total knee arthroplasty. J Arthroplast 19:459–463 28. VanKampen A, Huiskes R (1990) The 3-dimensional tracking pattern of the human patella. J Orthop Res 8:372–382 29. Wiles AD, Thompson DG, Frantz DD (2004) Accuracy assessment and interpretation for optical tracking systems. Med Imaging, Proc. 5367, visualization, image-guided procedures, p 12 30. Worland RL, Jessup DE, Johnson GVV, Alemparte JA, Tanaka S, Rex FS, Keenan J (2002) The effect of femoral component rotation and asymmetry in total knee replacements. Orthopedics 25:1045–1048

123