Accuracy of intramedullary alignment in total knee replacement

The accuracy of a system of intramedullary alignment using 6 mm rods was assessed in 100 patients undergoing total knee replacements. Post-operative, full length weight-bearing X-rays were used; the mechanical axis from head was used as the reference axis. The method of calculating the errors produced by flexion and rotation of the limb in relation to the X-ray beam is described, the mean deviation from the mechanical axis in 100 cases being 0.67° valgus with a standard deviation of 2.47°. The maximum error was 6.68° valgus and 4.62° varus. The purpose of this study is twofold, first to assess the accuracy of this system of intramedullary alignment and, second, to develop a method of correcting for apparent radiological misalignment using standard radiographic equipment.     

Evaluation of knee functional calibration with and without the effect of soft tissue artefact

Functional calibration methods were devised to improve repeatability and accuracy of the knee flexion–extension axis, which is used to define the medio-lateral axis of the femur coordinate system in gait analysis. Repeatability of functional calibration methods has been studied extensively in healthy individuals, but not accuracy in the absence of a benchmark knee axis. We captured bi-plane fluoroscopy data of the knee joint in 17 subjects with unilateral total knee arthroplasty during treadmill walking. The prosthesis provided a benchmark knee axis to evaluate the functional calibration methods. Stereo-photogrammetry data of thigh and shank marker clusters were captured simultaneously to investigate the effect of soft tissue artefact (STA). Three methods were tested, the Axis Transformation Technique (ATT) finds the best single fixed axis of rotation, 2DofKnee finds the axis that minimises knee varus–valgus and trajAJC finds the axis perpendicular to the trajectory, in the transverse plane of the femur, of a point located on the longitudinal axis of the tibia. Using fluoroscopy data, functional axes formed an angle of less than 2° in the transverse plane with the benchmark axis. True internal–external range of movement was correlated with decreased accuracy for ATT, while varus–valgus range of movement was correlated with decreased accuracy for 2DofKnee and trajAJC. STA had negative impact on accuracy and variability. Using stereo-photogrammetry data, the accuracy of 2DofKnee was 1.7°(SD: 5.1°), smaller than ATT 2.9°(SD: 5.1°) but not to trajAJC 1.7°(SD: 5.2°). Our results confirm that of previous studies, which utilised the femur condylar axis as reference.     

The effects of knee motion and external loading on the length of the anterior cruciate ligament (ACL): a kinematic study

A six-degrees-of-freedom mechanical linkage device was designed and used to study the unconstrained motion of ten intact human cadaver knees. The knees were subjected to externally applied varus and valgus (V-V) moments up to 14 N-m as well as anterior and posterior (A-P) loads up to 100 N. Tests were done at four knee flexion angles; 0, 30, 45, and 90 deg. Significant coupled axial tibial rotation was found, up to 21.0 deg for V-V loading (at 90 deg of flexion) and 14.2 deg for A-P loading (at 45 deg of flexion). Subsequently, the knees were dissected and the locations of the insertion sites to the femur and tibia for the anteromedial (AM), posterolateral (PL), and intermediate (IM) portions of the ACL were identified. The distances between the insertion sites for all external loading conditions were calculated. In the case when the external load was zero, the AM portion of the ACL lengthened with knee flexion, while the PL portion shortened and the intermediate (IM) portion did not change in length. With the application of 14 N-m valgus moment, the PL and IM portions of the ACL lengthened significantly more than the AM portion (p less than 0.001). With the application of 100 N anterior load, the AM portion lengthened slightly less than the PL portion, which lengthened slightly less than the IM portion (p less than 0.005). In general, the amount of lengthening of the three portions of the ACL during valgus and anterior loading was observed to increase with knee flexion angle (p less than 0.001).     

How changing the inversion/eversion foot angle affects the nondriving intersegmental knee moments and the relative activation of the vastii muscles in cycling

Nondriving intersegmental knee moment components (i.e., varus/valgus and internal/external axial moments) are thought to be primarily responsible for the etiology of overuse knee injuries such as patellofermoral pain syndrome in cycling because of their relationship to muscular imbalances. However the relationship between these moments and muscle activity has not been studied. Thus the four primary objectives of this study were to test whether manipulating the inversion/eversion foot angle alters the varus/valgus knee moment (Objective 1) and axial knee moment (Objective 2) and to determine whether activation patterns of the vastus medialis oblique (VMO), vastus lateralis (VL), and tensor fascia latae (TFL) were affected by changes in the varus/valgus (Objective 3) and axial knee moments (Objective 4). To fulfill these objectives, pedal loads and lower limb kinematic data were collected from 15 subjects who pedaled with five randomly assigned inversion/eversion angles: 10 deg and 5 deg everted and inverted and 0 deg (neutral). A previously described mathematical model was used to compute the nondriving intersegmental knee moments throughout the crank cycle. The excitations of the VMO, VL, and TFL muscles were measured with surface electromyography and the muscle activations were computed. On average, the 10-deg everted position decreased the peak varus moment by 55% and decreased the peak internal axial moment by 53% during the power stroke (crank cycle region where the knee moment is extensor). A correlation analysis revealed that the VMO/VL activation ratio increased significantly and the TFL activation decreased significantly as the varus moment decreased. For both the VMO/VL activation ratio and the TFL activation, a path analysis indicated that the varus/valgus moment was highly correlated to the axial moment but that the correlation between muscle activation and the varus moment was due primarily to the varus/valgus knee moment rather than the axial knee moment. The conclusion from these results is that everting the foot may be beneficial towards either preventing or ameliorating patellofemoral pain syndrome in cycling.     

Strategies of muscular support of varus and valgus isometric loads at the human knee

In this paper we studied how subjects activate their muscles in response to static varus and valgus loads at the knee. The muscles' contributions to the external moments were estimated using an EMG driven biomechanical model of the knee. The individual muscle activation and loading patterns were examined to identify the strategies that the nervous system uses to support varus and valgus knee moments. It was found that the (1) co-contraction of the hamstrings and quadriceps, and (2) activation of the gracilis and tensor fascia lata increased with the increasing magnitude of the varus and valgus moments. These 2 activation patterns provided positive support of valgus and varus loads at the knee The sartorius appears to be activated to provide positive support of valgus loads at the knee, whereas during varus moments this muscle increases the varus load on the knee, i.e. provides negative support. Generally, the hamstrings and quadriceps co-contraction contributed to most of the muscular support of the varus and valgus moments. In addition, co-contraction supported 11-14% of the external moment in pure varus and pure valgus respectively. It appears that there are activation strategies with the specific purpose to support varus and valgus moments, albeit small, which suggest dual goals of the neuromotor system during the support of varus and valgus moments.     

The mechanical consequences of dynamic frontal plane limb alignment for non-contact ACL injury

This study investigated the mechanical consequences of differences in dynamic frontal plane alignment of the support limb and the influence of anticipatory muscle activation at the hip and ankle on reducing the potential for non-contact ACL injury during single-limb landing. A frontal plane, three-link passive dynamic model was used to estimate an ACL non-contact injury threshold. This threshold was defined as the maximum axial force that the knee could sustain before the joint opened 8 degrees either medially or laterally, which was deemed sufficient to cause injury. The limb alignment and hip and ankle muscle contractions were varied to determine their effects on the ACL injury threshold. Valgus or varus alignment reduced the injury threshold compared to neutral alignment, but increasing the anticipatory contraction of hip abduction and adduction muscle groups increased the injury threshold. Increasing anticipatory ankle inversion/eversion muscle contraction had no effect. This study provides a mechanical rationale for the conclusion that a neutral limb alignment (compared to valgus or varus) during landing and increasing hip muscle contraction (abductors/adductors) prior to landing can reduce the possibility of ACL rupture through a valgus or varus opening mechanism.     

Non-driving intersegmental knee moments in cycling computed using a model that includes three-dimensional kinematics of the shank/foot and the effect of simplifying assumptions

Assessing the importance of non-driving intersegmental knee moments (i.e. varus/valgus and internal/external axial moments) on over-use knee injuries in cycling requires the use of a three-dimensional (3-D) model to compute these loads. The objectives of this study were: (1) to develop a complete, 3-D model of the lower limb to calculate the 3-D knee loads during pedaling for a sample of the competitive cycling population, and (2) to examine the effects of simplifying assumptions on the calculations of the non-driving knee moments. The non-driving knee moments were computed using a complete 3-D model that allowed three rotational degrees of freedom at the knee joint, included the 3-D inertial loads of the shank/foot, and computed knee loads in a shank-fixed coordinate system. All input data, which included the 3-D segment kinematics and the six pedal load components, were collected from the right limb of 15 competitive cyclists while pedaling at 225 W and 90 rpm. On average, the peak varus and internal axial moments of 7.8 and 1.5 N m respectively occurred during the power stroke whereas the peak valgus and external axial moments of 8.1 and 2.5 N m respectively occurred during the recovery stroke. However, the non-driving knee moments were highly variable between subjects; the coefficients of variability in the peak values ranged from 38.7% to 72.6%. When it was assumed that the inertial loads of the shank/foot for motion out of the sagittal plane were zero, the root-mean-squared difference (RMSD) in the non-driving knee moments relative to those for the complete model was 12% of the peak varus/valgus moment and 25% of the peak axial moment. When it was also assumed that the knee joint was revolute with the flexion/extension axis perpendicular to the sagittal plane, the RMSD increased to 24% of the peak varus/valgus moment and 204% of the peak axial moment. Thus, the 3-D orientation of the shank segment has a major affect on the computation of the non-driving knee moments, while the inertial contributions to these loads for motions out of the sagittal plane are less important.     

Failure locus of the anterior cruciate ligament: 3D finite element analysis

Anterior cruciate ligament (ACL) disruption is a common injury that is detrimental to an athlete's quality of life. Determining the mechanisms that cause ACL injury is important in order to develop proper interventions. A failure locus defined as various combinations of loadings and movements, internal/external rotation of femur and valgus and varus moments at a 25^o knee flexion angle leading to ACL failure was obtained. The results indicated that varus and valgus movements were more dominant to the ACL injury than femoral rotation. Also, Von Mises stress in the lateral tibial cartilage during the valgus ACL injury mechanism was 83% greater than that of the medial cartilage during the varus mechanism of ACL injury. The results of this study could be used to develop training programmes focused on the avoidance of the described combination of movements which may lead to ACL injury.     

Three-dimensional knee joint loading during seated cycling.

The hypothesis which motivated the work reported in this article was that neglecting pure moments developed between the foot and pedal during cycling leads to a substantial error in computing axial and varus/valgus moments at the knee. To test this hypothesis, a mathematical procedure was developed for computing the three-dimensional knee loads using three-dimensional pedal forces and moments. In addition to data from a six-load-component pedal dynamometer, the model used pedal position and orientation and knee position in the frontal plane to determine the knee joint loads. Experimental data were collected from the right leg of 11 male subjects during steady-state cycling at 90 rpm and 225 W. The mean peak varus knee moment calculated was 15.3 N m and the mean peak valgus knee moment was 11.2 N m. Neglecting the pedal moment about the anterior/posterior axis resulted in an average absolute error of 2.6 N m and a maximum absolute error of 4.0 N m in the varus/valgus knee moment. The mean peak internal and external axial knee moments were 2.8 N m and 2.3 N m, respectively. The average and maximum absolute errors in the axial knee moment for not including the moment about an axis normal to the pedal were found to be 2.6 N m and 5.0 N m, respectively. The results strongly support the use of three-dimensional pedal loads in the computation of knee joint moments out of the sagittal plane.     

Failure locus of the anterior cruciate ligament: 3D finite element analysis

Anterior cruciate ligament (ACL) disruption is a common injury that is detrimental to an athlete's quality of life. Determining the mechanisms that cause ACL injury is important in order to develop proper interventions. A failure locus defined as various combinations of loadings and movements, internal/external rotation of femur and valgus and varus moments at a 25(o) knee flexion angle leading to ACL failure was obtained. The results indicated that varus and valgus movements were more dominant to the ACL injury than femoral rotation. Also, Von Mises stress in the lateral tibial cartilage during the valgus ACL injury mechanism was 83% greater than that of the medial cartilage during the varus mechanism of ACL injury. The results of this study could be used to develop training programmes focused on the avoidance of the described combination of movements which may lead to ACL injury.     

The effect of lower-limb anatomy on knee loads during seated cycling.

Overuse knee joint injuries are the primary injuries to cyclists. Overuse injuries have been intuitively linked to the anatomic structure of the foot because external loads are applied to the foot in cycling. Thus, the structure and function of the foot should dictate in part how the loads are transmitted to the knee joint. Therefore, it was hypothesized that patterns in knee loads are related to the anatomic structure of the foot. To test this hypothesis, peak knee loads (dependent variables) were related to anatomical variables (independent variables) through statistical analyses. This required first the detailed evaluation (i.e. measurement) of the anatomical structure of the foot and leg for 23 subjects. Next, three-dimensional knee joint loads were determined for a standardized riding condition. The results of the statistical analyses indicated that a group of cyclists with the most extreme inversion of the forefoot relative to the transverse plane developed significantly greater average posterior knee force and extensive knee moment. In addition, a number of anatomical variables significantly accounted for the variability in peak values of the posterior force, the extensive moment, the varus/valgus moment and the external axial moment. Based on these results, the hypothesis is accepted.     

The effect of valgus/varus malalignment on load distribution in total knee replacements

Valgus or varus malpositioning of the tibial component of a total knee implant may cause increased propensity for loosening or implant wear and eventually may lead to revision surgery. The aim of this study was to determine the effect of valgus/varus malalignment on tibio-femoral mechanics during surgical trial reduction and simulated gait loading. In seven cadaver legs, posterior cruciate sparing total knee replacements were implanted and tibial inserts representing a neutral alignment and 3 degrees and 5 degrees varus and valgus alignments were sequentially inserted. Each knee with each insert was loaded in a manner representative of a trial reduction performed during knee surgery and loaded in a physiological knee simulator. Simulated gait performed on the simulator demonstrated that internal/external and adduction/abduction rotations showed statistical changes with some of the angled inserts at different points in the walking cycle. Neither medial/lateral nor anterior/posterior translations changed statistically during simulated walking. The pressure distribution and total load in the medial and lateral compartments of the tibial component changed significantly with as little as a 3 degrees variation in angulation when loaded in a manner representative of a trial reduction or with a knee simulator. These results support the need for precise surgical reconstruction of the mechanical axis of the knee and proper alignment of the tibial component. These results further demonstrate that tibial contact pressures measured during a trial reduction method may be predictive of contact mechanics at the higher loading seen in the knee simulator.     

Locomotion in the quail (Coturnix japonica): the kinematics of walking and increasing speed

Hindlimb segmental kinematics and stride characteristics are quantified in several quail locomoting on a treadmill over a six-fold increase in speed. These data are used to describe the kinematics of a walking stride and to identify which limb elements are used to change stride features as speed increases. In quail, the femur does not move during locomotion and the tarsometatarsus-phalangeal joint is a major moving joint; thus, quail have lost the most proximal moving joint and added one distally. The tibiotarsus and tarsometatarsus act together as a fixed strut swinging from the knee during stance phase (the ankle angle remains constant at a given speed) and the tarsometatarsus-phalangeal joint appears to have a major role in increasing limb length during the propulsive phase of the stride. Speed is increased with greater knee extension and by lengthening the tibiotarsus/tarsometatarsus via increased ankle extension at greater speeds. Because the femur is not moved and three distal elements are, quail move the limb segments through a stride and increase speed in a way fundamentally different from other nonavian vertebrates. However, the three moving joints in quail (the knee, ankle, and tarsometatarsophangeal joint) have strikingly similar kinematics to the analogous moving joints (the hip, knee, and ankle) in other vertebrates. Comparisons to other vertebrates indicate that birds appear to have two modes of limb function (three- and four-segment modes) that vary with speed and locomotory habits.     

Cross-Talk Correction Method for Knee Kinematics in Gait Analysis Using Principal Component Analysis (PCA): A New Proposal

Background

In 3D gait analysis, the knee joint is usually described by the Eulerian way. It consists in breaking down the motion between the articulating bones of the knee into three rotations around three axes: flexion/extension, abduction/adduction and internal/external rotation. However, the definition of these axes is prone to error, such as the “cross-talk” effect, due to difficult positioning of anatomical landmarks. This paper proposes a correction method, principal component analysis (PCA), based on an objective kinematic criterion for standardization, in order to improve knee joint kinematic analysis.

Methods

The method was applied to the 3D gait data of two different groups (twenty healthy subjects and four with knee osteoarthritis). Then, this method was evaluated with respect to three main criteria: (1) the deletion of knee joint angle cross-talk (2) the reduction of variance in the varus/valgus kinematic profile (3) the posture trial varus/valgus deformation matching the X-ray value for patients with knee osteoarthritis. The effect of the correction method was tested statistically on variabilities and cross-talk during gait.

Results

Cross-talk was lower (p<0.05) after correction (the correlation between the flexion-extension and varus-valgus kinematic profiles being annihilated). Additionally, the variance in the kinematic profile for knee varus/valgus and knee flexion/extension was found to be lower and higher (p<0.05), respectively, after correction for both the left and right side. Moreover, after correction, the posture trial varus/valgus angles were much closer to x-ray grading.

Conclusion

The results show that the PCA correction applied to the knee joint eliminates the cross-talk effect, and does not alter the radiological varus/valgus deformation for patients with knee osteoarthritis. These findings suggest that the proposed correction method produces new rotational axes that better fit true knee motion.     

Rotational flexibility of the human knee due to varus/valgus and axial moments in vivo

Knee ligamentous injuries persist in the sport of Alpine skiing. To better understand the load mechanisms which lead to injury, pure varus/valgus and pure axial moments were applied both singly and in combination to the right knees of six human test subjects. The corresponding relative knee rotations in three degrees of freedom were measured. Knee flexion angles for each test subject were 15 and 60 degrees for the individual moments and 60 degrees for the combination moments. For both knee flexion angles the hip flexion angle was 0 degrees. Leg muscles were quiescent and axial force was minimal during all tests. Tables of data include sample statistics for each of four flexibility parameters in each loading direction. Data were analyzed statistically to test for significant differences in flexibility parameters between the test conditions. In flexing the knee from 15 to 60 degrees, the resulting knee rotations under single moments depended upon flexion angle with varus, valgus, and internal rotations increasing significantly. Also, rotations were different depending on load direction; varus rotation was significantly different and greater than valgus rotation at both flexion angles. Also external rotation was significantly different and greater than internal at 15 degrees flexion, but not at 60 degrees flexion. Coupled rotations under single moments were also observed. Applying pure varus/valgus moments resulted in coupled external/internal rotations which were inconsistent and hence not significant. Applying pure axial moments resulted in consistent and hence significant varus/valgus rotations; an external axial moment induced varus rotation and an internal axial moment induced valgus rotation. For combination moments, varus/valgus rotations decreased significantly from those rotations at similar load levels in the single moment studies. Also, a varus moment significantly increased external rotation and a valgus moment significantly decreased internal rotation. These differences indicate significant interaction between corresponding load combinations. These results suggest that load interaction is a potentially important phenomenon in knee injury mechanics.     

Effects of femoral component placement on the balancing of a total knee at surgery

Misalignment and soft-tissue imbalance in total knee arthroplasty (TKA) can cause discomfort, pain, inadequate motion and instability that may require revision surgery. Balancing can be defined as equal collateral ligament tensions or equal medial and lateral compartmental forces during the flexion range. Our goal was to study the effects on balancing of linear femoral component misplacements (proximal, distal, anterior, posterior); and different component rotations in mechanical alignment compared to kinematic alignment throughout the flexion path. A test rig was constructed such that the position of a standard femoral component could be adjusted to simulate the linear and rotational positions. With the knee in neutral reference values of the collateral tensions were adjusted to give anatomic contact force patterns, measured with an instrumented tibial trial. The deviations in the forces for each femoral component position were then determined. Compartmental forces were significantly influenced by 2 mm linear errors in the femoral component placement. However, the errors were least for a distal error, equivalent to undercutting the distal femur. The largest errors mainly increase the lateral condyle force, occurred for proximal and posterior component errors. There were only small contact force differences between kinematic and mechanical alignment. Based on these results, surgeons should avoid overcutting the distal femur and undercutting the posterior femur. However, the 2–3 degrees varus slope of the joint line as in kinematic alignment did not have much effect on balancing, so mechanical or kinematic alignment were equivalent.     

Effect of muscular activity on valgus/varus laxity and stiffness of the knee

Quantitative changes in valgus/varus knee stability with different levels of muscular activity were determined for five subjects. A specially designed machine was used to measure resistance to angulation in the frontal plane. This device held the thigh stationary, the knee straight, an cycled the leg from side to side at a constant rate between present moment limits. Resistance to this forced valgus/varus motion was measured simultaneously with torque about the knee in the sagittal plane. Muscle activity was monitored by electromyography (EMG). Direct comparison of moment-rotation characteristics allowed changes in stability to be quantified as a function of extension and flexion torque. Extension torques less than 20% of the maximum increased varus stability more than valgus stability. Flexion torques of the same relative magnitude increased valgus stability more than varus stability. Comparison with the literature suggested that prevention of opening of the lateral side of the joint under varus loading was responsible for increased varus stability with increasing torque, both with extension and flexion torques.     

Skeletal allometry and interlimb scaling patterns in mustelid carnivorans

To address the effects of an evolutionary increase in body size on long bone skeletal allometry, scaling patterns relating body mass, bone length, limb length, midshaft diameters, and cross-sectional properties of the humerus and femur were analyzed for four species of scansorial mustelids. Humeral and, to a lesser extent, femoral allometry is consistent with expectations of elastic similarity: bone and limb length scale with negative allometry on body mass while bone robusticity (cross-sectional parameters against bone length) scales with strong positive allometry. Differences between fore- and hindlimb scaling patterns, however, are observed, with size-dependent increases in forelimb length and humeral strength and robusticity exceeding those of the hindlimb and femur. It is hypothesized that this greater fore- than hindlimb lengthening results in postural modifications that serve to straighten the hindlimb of larger bodied scansorial mustelids relative to smaller mustelids. Straightening of hindlimb joints would more precisely align the long axis of the femur with peak (vertical) ground reaction forces, thereby accounting for the reduction in relative bending stresses acting on the femur compared to the humerus. J. Morphol. 235:121–134, 1998. © 1998 Wiley-Liss, Inc.     

The Validity of a New Low-Dose Stereoradiography System to Perform 2D and 3D Knee Prosthetic Alignment Measurements

Introduction

The EOS stereoradiography system has shown to provide reliable varus/valgus (VV) measurements of the lower limb in 2D (VV2D) and 3D (VV3D) after total knee arthroplasty (TKA). Validity of these measurements has not been investigated yet, therefore the purpose of this study was to determine validity of EOS VV2D and VV3D.

Methods

EOS images were made of a lower limb phantom containing a knee prosthesis, while varying VV angle from 15° varus to 15° valgus and flexion angle from 0° to 20°, and changing rotation from 20° internal to 20° external rotation. Differences between the actual VV position of the lower limb phantom and its position as measured on EOS 2D and 3D images were investigated.

Results

Rotation, flexion or VV angle alone had no major impact on VV2D or VV3D. Combination of VV angle and rotation with full extension did not show major differences in VV2D measurements either. Combination of flexion and rotation with a neutral VV angle showed variation of up to 7.4° for VV2D; maximum variation for VV3D was only 1.5°. A combination of the three variables showed an even greater distortion of VV2D, while VV3D stayed relatively constant. Maximum measurement difference between preset VV angle and VV2D was 9.8°, while the difference with VV3D was only 1.9°. The largest differences between the preset VV angle and VV2D were found when installing the leg in extreme angles, for example 15° valgus, 20° flexion and 20° internal rotation.

Conclusions

After TKA, EOS VV3D were more valid than VV2D, indicating that 3D measurements compensate for malpositioning during acquisition. Caution is warranted when measuring VV angle on a conventional radiograph of a knee with a flexion contracture, varus or valgus angle and/or rotation of the knee joint during acquisition.     

A method to quantify alteration of knee kinematics caused by changes of TKR positioning

Few in-vitro studies have investigated changes in kinematics caused by total knee replacement (TKR) implantation. The advent of surgical navigation systems allows implant position to be measured accurately and the effects of alteration of TKR position and alignment investigated. A test rig and protocol were developed to compare the kinematics of TKR-implanted knees for different femoral component positions. The TKR was implanted and the component positions documented using a navigation system. The quadriceps was tensed and the knees were flexed and extended manually. Torques and drawer forces were applied to the tibia during knee flexion–extension, while recording the kinematics with the navigation system. The implant was removed and replaced on an intramedullary fixation that allowed proximal–distal, and internal–external rotation of the femoral component without conducting a repeated arthrotomy on the knee. The implant was repositioned using the navigation system to reproduce the previously achieved normally navigated position and the kinematics were recorded again. The recorded kinematics of the knee were not significantly different between both normal implantation and intramedullary remounting for tibial internal–external rotation, varus–valgus angulation, or posterior drawer, at any angle of knee flexion examined. Anterior drawer was increased approximately 2.5 mm across the range 20–35° knee flexion (p<0.05), but was otherwise not significantly different. This method of navigating implant components and of moving them within the closed knee (thus avoiding artefactual effects of repeated soft tissue manipulations) can now be used to quantify the effect on kinematics of alteration of the position of the femoral component.