There are concerns that highly crosslinked polyethylenes are not appropriate for knee implants. One concern is insufficient attachment strength of the insert to the tibial baseplate. In this study, the Natural Knee II Durasul (Sulzer Orthopedics Inc.) snap feature was optimized and then tested to evaluate if the locking mechanism withstood in vivo forces.

Initial testing showed that the anterior snap did not always fully engage if the conventional polyethylene design was employed. Therefore, a slight modification was made to the anterior face of the anterior snap feature. Subsequently, full engagement was consistently achieved.

The optimal snaplock geometry was evaluated using size 3, 22mm thick Durasul inserts in a peel-out test. Thick inserts were employed for a worst-case scenario (greatest lever arm). The modification employed resulted in a doubling of the attachment strength.

Once the optimal snap was defined, peel-out and constraint testing were conducted to determine in vivo performance characteristics the insert/baseplate attachment strength. The attachment strength of the Natural-Knee II Durasul tibial insert exceeds the maximum shear forces at the knee reported by Greenwald et al., even without an applied compressive load that would be present physiologically and would increase the attachment strength. This locking mechanism is stronger than clinically successful implants

To further verify the design, a 20° shear fatigue test was conducted on 22mm Ultracongruent inserts, again, a worst-case scenario. This study evaluated the migration of the polyethylene by monitoring anterior displacement of the femur and posterior lift-off under extreme physiological loads. All knee assemblies survived 107 fatigue cycles with no adverse effects. All inserts remained firmly attached to their respective baseplates. No polyethylene cracking was detected at the tabs of the insert or in the body of the inserts. This study shows that a successful locking mechanism can be made with the Durasul material.

Although complications associated with patello-femoral (PF) joint account for up to 50% of total knee replacement (TKR) revision procedures (Lee), the PF joint has been overlooked in wear simulations. The goal of this study was to develop an in vitro model to simulate patella wear in TKR’s. This report describes the concepts of an in vitro model for normal gait and the preliminary results of experimental validation.

The primary consideration in the development of the current model was modeling of the in vivo kinetics and kinematics. Since the in vivo kinetics are not well documented, the current model adapted a PF joint force pattern of gait measured one year postoperatively in a telemetric distal femoral replacement (Taylor et al). The maximum force was increased from 571N to 1780N (2.5xBody Weight) to compensate for muscle deficiency and to better reflect a maximum load representative of the in vivo situation. In vivo kinematics were adopted from measurements of Lafortune. Only the PF flexion was included in the model as a simplification of the complex patella motions. The phase relationship between the kinematic and kinetic waveforms was adjusted to replicate the in vivo situation. A 6-station knee simulator carried out the experimental validation with a test frequency of 1.5Hz. The test was intended to run for 5 million cycles, with CMM wear measurements (Muratoglu et al.) taken every million cycles. The preliminary measurements showed wear patterns in the tested patellae similar to retrieved patellae. Currently there are no standards for wear testing the PF joint. The current in vitro wear model presents a useful tool to critically assess the PF joint during gait. Future work should incorporate testing for adverse loading conditions, such as PF mal-alignment, rising from a chair or deep knee flexion.

A high proportion of complications following TKR occur at the patellofemoral articulation secondary to delami-nation and adhesive/abrasive wear. Electron beam cross-linking and melting has been shown to substantially reduce delamination and adhesive/abrasive wear in polyethylene tibial inserts. A series of in-vitro patella wear and fatigue tests were developed to explore the benefits of this material at the patellofemoral articulation.

Patellae (NKII, Sulzer Orthopedics, Inc., Austin, TX) were tested on an AMTI (Watertown, MA) knee simulator articulating against the trochlear grove of the femoral component. The simulator controlled flexion/ extension and patellofemoral contact force. Each test included patellae manufactured from conventional and electron beam crosslinked and melted polyethylene. Three different simulations were created: i) normal gait (5 million cycles) with optimal component alignment, ii) stair climbing (2 million cycles) with optimal component alignment, iii) stair climbing (2 million cycles) with 4° of femoral component internal rotation to simulate a component malalignment condition. In the last two simulations all patellae were artificially aged for 35 days in 80°C air to simulate one aspect of the long term oxidative state of each material.

In normal gait, the unaged conventional and highly cross-linked materials demonstrated similar behaviour. In stair climbing with optimal component alignment, the aged conventional patellae developed cracks by 2 million cycles. In stair climbing with component malalign-ment the aged conventional patellae developed cracks and delamination by 1 million cycles. None of the highly cross-linked components showed cracks or delamination. These results demonstrate the potential advantage of highly cross-linked polyethylene for the patella.