INTRODUCTION

In patients with neural disorders such as cerebral palsy, three-dimensional marker-based motion analysis has evolved to become a well standardized procedure with a large impact on the clinical decision-making process. On the other hand, in knee arthroplasty research, motion analysis has been little used as a standard tool for objective evaluation of knee joint function. Furthermore, in the available literature, applied methodologies are diverse, resulting in inconsistent findings [1]. Therefore we developed and evaluated a new motion analysis framework to enable standardized quantitative assessment of knee joint function.

Computer-assisted techniques in total knee replacement (TKR) have been introduced to improve bone cuts execution and relevant prosthesis components positioning. Although these have resulted in good surgical outcomes when compared to the conventional TKR technique, the surgical time increase and the use of additional invasive devices remain still critical. In order to cope with these issues, a new technology in TKR has been introduced also for positioning prosthetic components according to the natural lower-limb alignment. This technique is based on custom-fit cutting block derived from patient-specific lower-limb scan acquisition. The purpose of this study is to assess the accuracy of the custom-fit technology by means of a knee surgical navigation system, here used only as measurement system, and post-operative radiographic evaluations. Particularly, the performances of two different custom-fit cutting blocks realized from as many scan acquisitions have been here reported.

Thirty patients affected by primary knee osteoarthritis were enrolled in this study. Fifteen patients were implanted with GMK® (Medacta-International, Castel San Pietro, CH) and as many patients with Journey® (Smith&Nephew, London, UK). Both TKR designs were implanted by using custom-fit blocks for bone cut executions provided by the same TKR manufacturers according to a pre-operative web planning approved by the surgeon. Particularly, the cutting block for the former design was built from CT scan acquisition of the hip, knee and ankle, whereas that for the latter design from MRI scans acquisition of the knee and X-ray lower-limb overview. A knee surgical navigation system (Stryker®-Leibinger, Freiburg, Germany) was used for recording intra-operative alignment of bone cuts as performed by means of the custom-fit cutting blocks and relevant component positioning. Prosthetic components alignments were also assessed post-operatively on X-ray images according to a shape-matching technique. The accuracy of the custom-fit blocks was evaluated through the comparison between pre-operative planning, and intra/post-operative data. Discrepancies above 3° and millimeters were considered as outliers.

Within the patient cohort, nine cases were fully analyzed at the moment and here reported. Over them and except for one case, the discrepancy between pre-operative planned femoral/tibial resection level on the frontal plane and the corresponding measured intra-operatively was within 3 mm, being 5 mm in the worse case. Two outliers were observed for the corresponding femoral/tibial cut rotational alignment. Particularly, in one patient, the discrepancy in femoral cut alignment was of 8° in flexion and 6° in external rotation; in another patient this was of 4° in extension and 4° in external rotation in the femoral and tibial cut alignment, respectively. Post-operative radiographs evaluations for the final prosthetic components revealed that femoral/tibial alignment were within 3° in all cases, except for those patients that were already outliers.

These preliminary results reveal the efficacy of the custom-fit cutting block for TKR. These were generally fitted properly and final prosthetic components were accurately placed, although some discrepancies were observed. This new technology seems to be a valid alternative to conventional and computer-assisted techniques. More consistent conclusions can be deduced after final evaluation of all patients.

A linkage-based mathematical model was used to design a ligament-compatible prosthesis to keep certain ligament fibres isometric during passive motion. The sagittal plane talar component radius is about 50% longer than that of the normal talus, the tibial component is spherically convex. A fully conforming meniscal bearing is interposed between them. Experiments in cadaver specimens confirmed the mathematical prediction that the bearing moves forwards on both metal components during dorsi-flexion and backwards during plantar flexion.

Between July 2003 and July 2008, the prosthesis was implanted into 250 patients at nine hospitals in Northern Italy. By November 2007, 158 in 156 patients were seen at least six months post-operatively. Mean age was 60.5 years. The diagnosis was post-traumatic osteoarthritis in 127, primary osteoarthritis in 17, rheumatoid arthritis in 10.

The mean follow-up was 32.5 months. The pre-operative AOFAS score of 36.2 rose to 75.9, 79.3, 77.9, and 79.0 respectively at 12, 24, 36, 48 months. Dorsi-flexion increased from 0.1° to 9.7°, plantarflexion from 15.1° to 24.6°. In 30 patients at one hospital, the range of postoperative motion, 14° – 53°, was significantly correlated to the range of bearing movement on the tibial component, 2mm–11mm, measured radiologically, (r2 = 0.37, p < 0.0005).

By December 2007, 2 revision operations had been performed at 24 months, one for unexplained pain not relieved by a successful arthrodesis, one in a patient with Charcot-Marie-Tooth disease. There were no device-related revisions (loosening, fracture, dislocation). The Kaplan-Meier survival rate (component-removal as end-point) at 4 years was 96% (Confidence interval 90–100%).

Early clinical results have demonstrated safety and efficacy. The survival rate at four years compares well with multi-centre 5-year rates published by the Swedish (531 cases, survival 78%), Norwegian (257, 89%) and New Zealand (202, 86%) registries.

A new design of total ankle replacement was developed. According to extensive prior research, the design features a spherical convex tibial component, a talar component with radius of curvature in the sagittal plane longer than that of the natural talus, and a meniscal component fully conforming to these two. The shapes of the tibial and talar components are compatible with a physiologic ankle mobility and with the natural role of the ligaments.

Within an eight-centre clinical trial, 114 patients were implanted in the period July 2003 – September 2006, with mean age 62.2 years (range 29 – 82). The AOFAS clinical score systems and standard radiographic assessment were used to assess patient outcome, here reported only for those 75 patients with follow-up longer than 6 months.

Intra-operatively, the components maintained complete congruence at the two articulating surfaces of the meniscal bearing over the entire motion arc, associated to a considerable anterior motion in dorsiflexion and posterior motion in plantarflexion of the meniscal-bearing, as predicted by the previous mathematical models. Mean 10.0 and 23.5 degrees respectively of dorsi- and plantar-flexion were measured immediately after implantation, for a mean additional range of motion of 19.2, which was maintained at follow-ups. Radiographs showed good alignment and no signs of evolutive radiolucency or loosening. The mean AOFAS score went from 40.8 pre-op to 66.2, 74.6 and 77.2 respectively at 3, 6 and 12 month follow-ups. One revision only was performed successfully three days after implantation because of a technical error.

In the score system utilized, Function and RoM sections scored better than any average previous total ankle result, Pain scored similarly. The satisfactory though preliminary observations from this novel design encourage continuation of the implantation, which is now extended over a few European countries. Instrumented gait and three-dimensional fluoroscopic analyses are in progress to quantify functional progresses.

Navigation-assisted surgery in total knee arthroplasty (TKA) is aimed at improving the accuracy with which prosthesis components are implanted in the bones, according to anatomical plane orientations. Traditional surgical techniques based on the identification of transepicondylar and intramedullary axes are replaced with those based on segmental anatomical frame definitions following anatomical landmark identification. These frames are offered on the screen to the surgeon to target in real time the alignment goal by adjusting position and orientation of the bone saw guides. However, immediately after sawing, final bone, and in case cement, preparation and component implantation is necessarily a series of actions performed manually by the surgeon. In the current study, we wanted to compare intra-operatively the final component alignments with the corresponding at the original resection planes.

In this series, 50 Scorpio PS TKAs were analyzed. The navigation system used was the Stryker Knee Navigation System (Stryker-Navigation, Kalamazoo, USA). An ‘anatomical survey’ defined anatomical frames for the femur and tibia, based on relevant anatomical landmark identification, and provided target orientations for all the relevant bone cuts. These references were targeted in all three anatomical planes, and bone cuts were made accordingly. Corresponding alignments of the bone resection planes in the frontal, sagittal and transverse planes for the femur and in the frontal and sagittal planes for the tibia were recorded, with a 0.5° resolution. Then, component implantation was performed and alignments were measured again by means of an instrumented probe. Because of the shape of the prosthesis components, only the alignments in the frontal plane for the femur and in the frontal and sagittal planes for the tibia were recorded.

The difference between the alignment of the bone cuts and the alignment of the prosthesis components, in the frontal plane of the femur, and in the frontal and sagittal planes of the tibia was larger than 2° respectively in 8%, 6%, 10% of the patients.

The present study offers a figure for the different alignment between resection planes and final implanted components, necessarily the effect of the manual procedures implied in TKA for the final implantation of the components. Considering that 1° is the claimed achievable accuracy of the navigation systems, and that the correct alignment goal was achieved at the resection planes, these figures reveal that in up to 10% of the patients the benefit obtained by navigation can be lost by the manual procedures implied in component implantation. These differences in alignment put also concerns in the postoperative statistical comparison between conventional and navigated TKAs.

Patellar maltracking after total knee arthroplasy (TKA) introduces complications such as anterior knee pain and patellar subluxation, generally due to prosthetic component malallignment in both tibiofemoral (TF) and patellofemoral joints. It is still debated if it is necessary to resurface the patella, which would better adapt the patellar articular surface to the prosthetic femoral troclea with a prosthesis, but also result in possible bone fractures. In this study, an in-vitro analysis is presented in order to identify differences between intact and TKA patellar tracking with and without patellar resurfacing and to show how much the latter is similar to intact knee patellar tracking.

Three fresh-frozen amputated legs with knees free from anatomical defects and with intact joint capsule, collaterals and quadriceps tendon were analyzed using the Stryker knee navigation system (Kalamazoo, MI-USA). Landmark digitations were used to define anatomical frames for femur, tibia and patella. Manually driven TF flexions, from 0 to 140, were performed under conditions of no load and of 10 kg on the quadriceps, with intact knee and TKA with patella resurfaced and not. TF flex/extension, intra/extra rotation, ad/abduction were calculated according to a standard convention. Patellar flex/extension, medial/lateral tilt, rotation and shift were calculated according to a recently proposed articular convention.

Since more repeatable, results relative to trials under 10 kg are reported. Intact knee: 4 abduction; considerable intra rotation (from 16 to 4), followed by continuous extra rotation starting at 30 TF flexion; linear increase in patellar flexion (from 20 to 110); initial medial patellar rotation (from 12 to 8), followed by medial rotation starting at 60 TF flexion; initial lateral patellar tilt (from 4 lateral to 4 medial), followed by medial tilt starting at 70 TF flexion; initial 6 mm lateral patellar shifts from 0 to 80 TF flexion, followed by 4 mm medial shift. TKA knee: small differences in ad/abduction between intact and TKA knees, both with and without resurfaced patella; slight initial extra rotation, followed by continuous intra rotation starting at 20 TF flexion; linear increase in the flexion of the patella, both resurfaced and not, close to the that of the intact knee; patellar rotation more lateral than in the intact knee; patellar tilt without resurfaced patella closer to the intact knee one; 6 mm lateral patellar shift, likely accounted for the surgical technique.

Slightly more than TKA with resurfaced patella, TKA with non resurfaced patella flexes nearly like the intact knee. The closeness in values of patellar flexion and tilt represents a proof of the closeness in behavior of not resurfaced patella in TKA to the patella in the intact knee.

Tibial component loosening continues to be the most common mode of TKA failure. A debate persists on the dependence of mobilisation of this component on the equilibrium between mechanical load transfer and counterbalancing bone resistance. The aim of the present work is to study the in-vivo kinematics of TKA and to relate it with the degree of posterior slope with which the tibial component was implanted for two prosthesis designs with congruent polyethylene insert.

Twenty-three patients with osteoarthritis of the knee had TKA using a cemented prosthesis (OPTETRAK, Exactech). A cruciate retaining (CR, 10 knees) or a posterior stabilized (PS, 13 knees) implant was randomly assigned at operation. Standard pre- and post-operative antero-posterior and lateral roentgenograms of the knee were taken. Fluoroscopic analysis was performed after at least 18 and 7 months after surgery for the CR and the PS group, respectively. Patients performed stair ascending, chair rising-sitting and step up-down motor tasks. Articular contacts were assumed as the two points on the medial and lateral femoral prosthetic condyles closest to the tibial component base-plate. The spine-cam distance was calculated as the minimum distance between corresponding surfaces.

Only small differences in the position of the contacts over knee flexion angles were found among the motor tasks and between the two TKA designs. An overall posterior location of the tibio-femoral contact points was found at the medial and lateral compartments over all motor tasks, a little more pronounced for the PS patients. Statistically significant correlation over the three motor tasks analysed was found between posterior position of the tibio-femoral medial contact in maximum knee flexion and the post-operative tibial posterior slope. This is true for the PS and for the aggregated groups. Although no statistically significant, a general trend is observed of higher degree of flexion at which the cam contacts the spine as the post-operative posterior slopes increases: a 35 higher knee flexion angle for a tibial component implanted with a 5 of posterior slope. Generally, even when the correlations were statistically significant the correlation coefficients were always lower than 0.4.

The present work reports combined measurements of post-operative posterior slope and full in-vivo relative motion of the components in both CR and PS TKAs. General trends were found between posterior slope of the tibial component and positions of the tibio-femoral contacts, but a statistically significant correlation was found only for the tibio-femoral medial contact in maximum knee flexion in the PS and in the aggregated. General trends were found between posterior slope of the tibial component and degree of flexion at which the cam starts to be in contact with the spine. The nearly standard antero-posterior translation of the tibio-femoral contacts can be bigger in flatter polyethylene inserts.

Aims: Prior research has demonstrated that currently available total ankle implants fail to restore physiologic joint mobility. Most of the modern mobile-bearing designs that feature a flat tibial component and a talar component with anatomic curvature in the sagittal plane function non physiologically with the natural ligament apparatus. The aims of this investigation were a) to elucidate the natural relationship between ligaments and articular surfaces at the intact human ankle joint and b) to develop a new design of total ankle replacement able to replicate this relationship between the retained ligaments and the implanted prosthetic components.

Methods: Motion during passive flexion was analyzed in ten skeleto-ligamentous lower leg preparations including tibia, fibula, talus, calcaneus and intact ligaments. Geometry of ligament fiber arrangement and articular surface shapes was obtained with a 3D digitizer (FARO Technologies, Inc.). A sagittal four-bar linkage model was formulated as formed by the tibia/fibula and talus/ calcaneus rigid segments and by the calcaneofibular and tibiocalcaneal ligaments. To test the ability of possible new prostheses to reproduce the compatible mutual function between the articulating surfaces and the ligaments retained, non-conforming two-component and fully-conforming three-component designs were analyzed. A new total ankle replacement has been designed, prototypes manufactured and implanted in seven skeleto-ligamentous lower leg preparations, and motion was observed. A corresponding new prosthesis has been produced (Finsbury, UK), and implanted in four patients.

Results: The articular surfaces and the ligaments alone prescribed joint motion into a preferred single path of multiaxial rotation (one degree of unresisted freedom). Fibers within the calcaneofibular and tibiocalcaneal ligaments remained most isometric throughout the passive range. The four-bar linkage model well predicted the sagittal plane kinematics observed in corresponding experiments. A ligament-compatible, convex-tibia, fully-congruent, three-component prosthesis design showed the best features: complete congruence over the entire range of flexion together with an acceptable degree of entrapment of the meniscal bearing. Restoration of natural joint kinematics and ligament recruitment was observed in all replaced ankles.

Conclusions: The overall investigation is demonstrating that a profound knowledge of the changing geometry of the joint passive structures throughout the range of passive flexion (mobility) is mandatory for a successful design of joint replacements.

Aims:Understanding total knee replacement mechanics and their influence on patient mobility requires accurate analysis of both operated joint accurate kinematics and full body kinematics and kinetics. The main aim of this study is to perform these two analyses conjointly, as never been reported previously. An innovative graphic-based interface is also pursued aimed at supporting quantitative functional assessment of these patients during the execution of daily living motor tasks in a single synchronized view.

Methods: Three-dimensional fluoroscopic and gait analysis were carried out on eleven patients with PCL-retaining mobile bearing (Interax ISA, Stryker / How-medica / Ostetonics) and on ten posterior stabilized fixed bearing (Optetrak PS, Exactech) knee prostheses. Patients performed three trials of stair ascent twice on the same day: first in the radiology department for fluoroscopy acquisition and later in the Movement Analysis Laboratory, utilizing an identical staircase. Three-dimensional fluoroscopic analysis entails reconstruction of absolute and relative positions and orientations of the two metal components in space by analyzing series of fluoroscopic images of the operated knee and utilizing knowledge of the 3D cad models of these components. Conventional stereophotogrammetry and dynamometry were used to calculate kinematics and kinetics of the trunk, pelvis and of the major joints of the lower limb. An advanced computer-based interface was developed (MULTIMOD, EU-funded project: IST-2000-28377) to show together a) original video of the patient tasks, b) 3D graphical representation of bony segment motion, c) original fluoroscopic images, d) 3D reconstruction of prosthesis component relative motion, and e) graphical transverse plane representation of the contact areas at the base-plate of the replaced knee. All these were registered in space and synchronized in time.

Results: No significant statistical differences on clinical data were found between the two patient populations. Observations at the interface allowed distinct identification of the most critical phases of the task and of the most common compensatory mechanisms utilized by these patients. Statistically significant correlation was found between knee flexion at foot strike and the position of the mid-condylar contact points, and between maximum knee adduction moment and corresponding lateral trunk tilt.

Conclusions: A more complete and powerful assessment of the functional performances of different TKR designs is obtained by combining gait and fluoroscopic in-vivo analyses, which provide correlated and synergic quantitative information.

Only recently has the mobility of the ankle joint been elucidated. Sliding/rolling of the articular surfaces and slackening/tightening of the ligaments have been explained in terms of a mechanism guided by the isometric rotation of fibres within the calcaneofibular and tibiocalcaneal ligaments. The purpose of this investigation was to design a novel ankle prosthesis able to reproduce this physiological mobility.

A four-bar linkage computer-based model was used to calculate the shapes of talar components compatible with concave, flat and convex tibial components and appropriate fully congruous meniscal bearings. Three-component designs were analysed, and full congruence of the articular surfaces, appropriate entrapment of the meniscal bearing and isometry of the two ligaments were required.

A convex tibial component with 5 cm arc radius gave a 2 mm entrapment together with a 9.8 mm amount of tibial bone cut, while maintaining ligament elongation within 0.03 % of the original length. The physiological patterns of joint motion and ligament tensioning were replicated. The talar component slid backwards while rolling forwards during dorsiflexion. These movements were accommodated by the forward displacement of the meniscal bearing on the tibial surface under the control of the ligaments. The complementary surfaces provide complete congruence over the entire range of flexion, such that a large contact area is achieved in all positions.

To restore the physiological mobility at the ankle joint, not only should the components be designed to be compatible with original ligament pattern of tensioning, but also these should be mounted in the appropriate position. A suitable surgical technique was devised and relevant instrumentation was manufactured. Five below-knee amputated specimens replaced with corresponding prototype components showed good agreement with the model predictions.

Current three-component designs using a flat tibial component and physiological talar shapes cannot be compatible with physiological ligament function.