Session II: Polyethylene Wear

Polyethylene wear has been the limiting factor to the long-term success of total knee arthroplasty from the time of its inception. Unlike the highly congruent ball and socket articulation in the hip, the geometry and articulation of the knee is complex. As engineers began to design implants for knee arthroplasty, they recognized from the outset that wear was likely to be a limiting factor in total knee arthroplasty because of the reduced contact area and high stresses on the polyethylene. Therefore, the early designs of the 1970s were single axis, highly congruent implants, designed to avoid excessive wear from high contact stresses. Triplaner motion was limited by the implants themselves, not by the soft tissues of the knee. Because stresses were concentrated primarily at the fixation interface, implant designs such as the polycentric (gunston) knee, the Geomedic (Howmedica, Rutherford, NJ), and hinged devices, had unacceptably high rates of aseptic loosening. ^6,7,9,13

Problems with loosening in the first generation of knee arthroplasty components limited interest in total knee arthroplasty except for among elderly patients and patients with low physical demands, such as those with rheumatoid arthritis. To address the issue of loosening, condylar designs were introduced in the late 1970s. These implants placed less stress on the fixation interface and allowed more freedom of motion between the knee components. Success with less congruent, condylar-shaped components, such as the Total Condylar Knee (Howmedica), increased the interest in total knee arthroplasty for patients with gonarthrosis ¹²; however, function was limited, primarily because of restricted knee flexion. With the concept of improving knee flexion by enhancing femoral rollback, manufacturers introduced the Insall-Burstein posterior-stabilized knee replacement (Zimmer, Warsaw, IN). Excellent results with improved knee flexion and excellent durability were achieved with this nonmodular implant. ^14,15

At this point in the evolution of knee implants, the results after total knee arthroplasty were so good that clinicians reported lower wear rates with total knee arthroplasties than with total hip arthroplasties. ¹⁴ Despite the recognized higher contact stresses in knees, these implants had numerous common features that contributed to low wear rates and rare wear-related problems such as osteolysis. During this time, the components were nonmodular, either all-polyethylene or compression molded one-piece components. The polyethylene thickness of the tibial component was 8-mm or greater and the implants were sterilized with ethylene oxide, a method that did not create free radicals and the potential for polyethylene oxidation. Because these components were moderately congruent and curved in the coronal and sagittal planes, contact stresses were acceptable.

Although clinical reports were comparable with the best reports with hip implants, retrieval studies were not as favorable. ⁸ Investigators reported greater wear rates with retrieved total knee components than with retrieved hip components. ⁵ In addition, the modes of wear were different. In the knee components, a common mode of wear included pitting and, in some instances, late onset delamination. This supported the hypothesis that wear of knee implants included fatigue wear because of contact stresses that exceeded the yield strength of polyethylene. Imbedded cement debris and abrasions in the polyethylene commonly were found in the retrieved knee components, confirming that surgical technique is more important for minimizing wear in total knee arthroplasty than in total hip arthroplasty.

By 1990, osteolysis had not presented as a clinical problem. Aseptic loosening continued to be the major cause of failure with the moderately congruent, condylar knee designs. To address failure by aseptic loosening, implants were made less congruent in the early 1980s to allow even more triplaner motion. These implants essentially were flat-on-flat in the coronal and the sagittal planes, and relied more on ligamentous support and ligament balancing to provide knee stability. Only the central eminence, or the nonarticulating central region of the tibial tray, provided resistance to subluxation. An advantage of such a design was that components of different sizes could be used on the femur and the tibia. A disadvantage with less congruent designs was that acceptable knee stability was more dependent on surgical technique.

To address the issue of aseptic loosening and improve fixation, metal-backing and modularity of the tibial component were introduced, but at the expense of thinner polyethylene. Metal-backing provided a means of augmenting the cementless tibial component fixation with cancellous screws to provide additional stability. The metal-backed tray design also created the opportunity to use polyethylene inserts of different thickness with varying degrees of constraint. The orthopaedic community readily accepted these design changes. However, from a wear standpoint, these changes were a compromise.