Tibial insert undersurface as a contributing source of polyethylene wear debris.

Sixty-seven ultrahigh molecular weight polyethylene tibial inserts from cementless total knee arthroplasties were retrieved at autopsy and revision surgery and analyzed for evidence of articular and nonarticular surface wear after a mean implantation time of 62.8 months (range, 4-131 months). Polyethylene cold flow and abrasive wear on the nonarticular insert surface (undersurface) were assigned a wear severity score (Grade 0-4). The severity of articular wear was assessed quantitatively and graded. Corresponding prerevision radiographs were evaluated for evidence of tibial metaphyseal osteolysis and osteolysis around tibial fixation screws. Exact nonparametric conditional inference methods were used to establish correlations between different variables and the occurrence of tibial metaphyseal osteolysis. Severe Grade 4 wear of the tibial insert undersurface was associated with tibial metaphyseal osteolysis or osteolysis around fixation screws. Time in situ statistically was related to Grade 4 undersurface wear and tibial metaphyseal osteolysis. The occurrence of tibial osteolysis was not related statistically to articular wear severity, insert thickness, or implant type. The main articulation between the femoral implant and ultrahigh molecular weight polyethylene insert has been assumed to be the primary source of polyethylene debris contributing to osteolysis and total knee arthroplasty implant failure. The undersurface of the insert is an additional source of polyethylene debris contributing to tibial metaphyseal osteolysis. To lessen polyethylene debris produced at this modular interface, the tibial implant locking mechanism should fix the insert firmly to the metal backing to decrease relative micromotion. Because motion between the insert and metal backing may be inevitable, the wear characteristics of the inner tray surface should be optimized to minimize wear debris production at this other articulation.

The tradeoffs associated with modular hip prostheses.

In an effort to gain greater insight into the tradeoffs associated with modular hip prostheses, 2 approaches were taken. A questionnaire was sent to each of the orthopaedic implant manufacturing companies asking specific questions regarding modular components, and a series of retrieved prostheses, both modular and nonmodular, were examined to determine the potential sources of problems associated with modular connections. The respondents to the questionnaire generally agreed that it was more expensive to produce modular prostheses due to the required tolerances at the modular connections, and that the increased flexibility provided by the modularity was important to surgical outcome. There was less consensus on whether inventories were reduced and little data to support any improvement in surgical outcome caused by modularity. The most frequent problems associated with modular connections were fretting and corrosion. Easily observable significant fretting occurred in 4% of 701 head/neck tapers. Corrosion was observed in > 30% of the mixed-alloy head/stem combinations, in < 10% of all-titanium-alloy modular components, and in < 6% of all-cobalt-alloy devices. In 1 series of retrieved modular femoral components (15 titanium alloy and 15 cobalt alloy) with both sets having approximately the same duration of implantation, 7% of the all-cobalt-alloy components had corrosion, whereas 33% of the mixed-alloy components had corrosion.

Loss of hydroxyapatite coating on retrieved, total hip components.

The goal of this study was to examine the early retrievals of hydroxyapatite-(HA) coated hip prostheses to assess evidence of osteoconductivity, resorption of HA, and the integrity of the HA/implant bond. Six retrieved HA-coated hip prostheses (3 femoral hip stems, 3 acetabular cups) were analyzed for the amount of bone ongrowth or ingrowth of the HA-coated surface and the extent to which the coating was still present after in vivo service. The examination of these six HA-coated prostheses indicates that HA appeared to be osteoconductive. There was evidence of debonding of HA from the smooth-surfaced femoral prosthesis, although that may have been a result of the extraction process. The five plasma-spray surfaced, HA-coated prostheses showed evidence of considerable loss of the HA coating at the time of receipt in the authors' laboratory, although it is was not possible to determine the cause of the loss of coating.

Failure of a well-fixed bone-ingrown titanium hip prosthesis.

A cementless titanium femoral stem was revised 5 years after implantation because of acute pain and progressive osteolysis. Substantial amounts of titanium and polyethylene wear debris were found in the surrounding tissues. Multiple sources of this debris were found as well as detachment of titanium fiber-mesh pads from the body of the femoral stem.

Interfaces between host and uncemented femoral hip prostheses.

The authors analyzed 200 uncemented hip prostheses that were retrieved. They determined the type and extent of host-implant interfaces, but found little correlation between subjective pain relief and the histology of the interfaces. Wear debris, bone resorption and looseness were related to pain, although even with these the relationship was not statistically significant.

Mechanisms of failure of modular prostheses.

The expectations of wear and longevity of total hip components are based in large part on Charnley's early work. The evolution of the total hip from the one-piece, all-polyethylene acetabular component and fixed-head femoral component to the myriad of parts that comprise many of today's total hip designs has brought with it an array of potential mechanisms for failure that were not present in the earlier design. The risk/benefit ratio of these new designs may need to be reevaluated based on the additional mechanisms for failure that they provide. One hundred eleven acetabular hip prostheses and 139 femoral prostheses, all of modular configuration, retrieved by surgeons in the field, and sent for histologic examination, were analyzed for this study. A number of component characteristics were found to be correlated to early failure. These included acetabular designs with thin polyethylene bearings, poor fixation of the polyethylene to the metal shell, and geometries that permitted a moment to be applied to the bearing insert, tending to cause it to rotate in the metal shell. Modular femoral components were observed to be susceptible to corrosion, with titanium-alloy stems mated to cast cobalt-alloy heads at greatest risk attributable to a galvanic effect. All modular connections of femoral and acetabular components are at risk for disassociation and fretting; therefore, clever design and precision machining are necessary to produce prostheses in which the benefits of modularity exceed the risks.

Corrosion between the components of modular femoral hip prostheses.

We studied the tapered interface between the head and the neck of 139 modular femoral components of hip prostheses which had been removed for a variety of reasons. In 91 the same alloy had been used for the head and the stem; none of them showed evidence of corrosion. In contrast, there was definite corrosion in 25 of the 48 prostheses in which the stem was of titanium alloy and the head of cobalt-chrome. This corrosion was time-dependent: no specimens were corroded after less than nine months in the body, but all which had been in place for more than 40 months were damaged. We discuss the factors which may influence the rate of these changes and present evidence that they were due to galvanically-accelerated crevice corrosion, which was undetected in previous laboratory testing of this type of prosthesis.

All-polyethylene patellar components are not the answer.

Recent reports of the high incidence of polyethylene failure of metal-backed patellar components have rekindled the interest in cemented, all-polyethylene designs. One hundred four retrieved patellar components of metal-backed and all-polyethylene designs were analyzed for wear. Additionally, patellofemoral contact stress as a function of flexion angle was measured for unused components using pressure-sensitive Fuji film. Significant wear (2+ on a 0-3 scale) was seen in 65% of metal-backed designs and in 78% of all-polyethylene components. Severe wear (3) was seen in 44% of all-polyethylene components and in 39% of metal-backed devices. The incidence of severe wear (3) of congruent designs was statistically significantly lower than that of the noncongruent designs. Contact stress analysis confirmed that dome-type geometries typically resulted in stresses that exceed the yield strength of the polyethylene, whereas the more congruent geometries generated significantly reduced stress levels.

The biomechanical problems of polyethylene as a bearing surface.

The metal backings of acetabular components can reduce the available polyethylene thickness, often to an alarming extent. This study indicated that the pitting and cracking of thin polyethylene surfaces have some similarities to tibial and patellar bearings and that creep-related deformation occurred more frequently in thin polyethylene components. Additionally, it appears that dimensional tolerances of the polyethylene inserts are difficult to maintain and may result in a nonuniform fit of both the femoral head into the component and the component into its own metal backing, which can lead to component separation. It is difficult to accurately measure the changes in material and mechanical properties of polyethylene over time. An additional ramification is that flaws, such as voids in the polyethylene, cannot be attributed to problems of bulk supply, fabrication, or postmanufacturing treatment. Orthopedic device manufacturers should keep samples of each lot of polyethylene used and provide components with serial numbers so that the source, composition, and properties of the original bulk material and material "as fabricated" can be documented. This would permit researchers studying revisions or postmortem samples to determine the changes in the polyethylene over time in vivo, thus improving the understanding of this crucial material. If manufacturers were to include dimensions and tolerances of the polyethylene inserts in their product literature, accurate measurements of wear of retrieved specimens may be possible.

Macroscopic and microscopic evidence of prosthetic fixation with porous-coated materials.

The host response to porous-coated prostheses appears favorable; there is little evidence of any adverse tissue response or significant osteoclastic activity except in grossly loose specimens. While the nature of retrieval specimens makes any statistical correlation problematic, some generalizations can be made. Femoral hip prostheses are most likely to present bone ingrowth along the lateral quadrant of their porous coating. The frequency of bone ingrowth of femoral components was nearly twice that of acetabular devices. Pore size, geometry, and porous-coating composition did not appear to influence the appearance of bone and fibrous tissue ingrowth. Direct bonding of bone to the uncoated portion of the prosthesis was rarely seen and occurred only in closest proximity to the porous-coated regions. Indications of pain and looseness are evidence that fibrous tissue ingrowth alone is not always sufficient to ensure stability. Additionally, some bone-ingrown prostheses were retrieved because of pain, which leads to the conclusion that local bone ingrowth cannot ensure a general freedom from pain, especially with partially coated prostheses. Bone and fibrous tissue response to the porous coatings generally consists of interdigitation, while the response to uncoated regions is fibrous tissue encapsulation. Burnishing the distal tips of many of the partially coated femoral prostheses is an indication of relative motion in that region, which may be a potential source of pain.