Particles generated at the non articulating surface (backside) of modular acetabular components have been implicated in the development of periprosthetic osteolysis after total hip arthroplasty. Several design changes have been introduced in modern uncemented acetabular cups in an attempt to reduce backside wear, including the use of so called “non-modular cups”. We compared the backside wear of retrieved cementless non-modular cups, with modular cups of first and second generation designs.

Nine retrieved non-modular cups (Implex) were match paired for time in situ, patient age and weight, with 9 retrieved Trilogy cups, 9 Harris-Galante 1, and 9 Harris-Galante 2. The average time in situ was 2.5 years (1 to 7). The backside was divided in quadrants and each rated utilizing a score with a value from 0 (absence of wear) to 3 (severe backside wear) for a total ranging from 0 to 12. The score was validated for intra and inter observer reproducibility.

Among 36 quadrants in the HG1 group there were 3 rated 1, 23 rated 2, and 10 rated 3. In the HG2 group, there was 1 quadrant rated 0, 16 rated 1, 14 rated 2, and 5 rated 3. In the Trilogy group, there were 6 quadrants rated 0, 27 rated 1, and 3 rated 2. In the Implex group, there were 15 quadrants rated 0, 21 rated 1. The average total backside wear score and 95%CI were 8.4 (7.6–9.3); 7.3 (5.5–9.1); 3.7 (3.2–4.1); and 2.3 (1.3–3.4) respectively. The HG1 and HG2 groups demonstrated similar backside wear scores (p=0.3). The HG1 and HG2 designs demonstrated significantly more backside wear than the Trilogy and Implex (p< 0.01). The differences between the Trilogy and the Implex were not significant in this cohort.

A comparison of the “in vivo” backside wear of first generation and modern acetabular cups has not been published to date. Despite the limitations imposed by the small sample studied, the presence of multiple screw holes in the HG retrievals, and sub-optimal matching for sex, height, and varied indication for revision, we detected significant reduction in the backside wear of modern modular and non-modular acetabular cups when compared to first generation modular designs.

Backside wear is generated at the non-articulating surfaces of modular acetabular cups. We compared the backside wear of retrieved liners from cementless non-modular and modular cups of first and second generation designs. We match paired for time in situ, patient age and weight, 9 retrieved Harris Galante type 1 liners, 9 Harris Galante type 2, 9 Trilogy, and 9 liners from a modern two-piece preassembled cup (Implex). The average time in situ was 2.5 years (1 to 7). The backside was divided in quadrants and each one examined under a 10x binocular loupe and rated with a score from 0 (absence of wear) to 3 (severe backside wear) for a total ranging from 0 to 12. Among 36 quadrants in the HG1 group there were 3 rated 1, 23 rated 2, and 10 rated 3. In the HG2 group, there was 1 quadrant rated 0, 16 rated 1, 14 rated 2, and 5 rated 3. In the Trilogy group, there were 6 quadrants rated 0, 27 rated 1, and 3 rated 2. In the Implex group, there were 15 quadrants rated 0, 21 rated 1. The average total backside wear score was 8.4; 7.3; 3.7; and 2.3 respectively. The HG cups demonstrated more severe backside wear than the Trilogy and Implex (HG1 vs Trilogy and HG1 vs Implex: p< 0.001; HG2 vs Trilogy and HG2 vs Implex: p< 0.02). There was a tendency towards less backside wear in the Implex cup when compared to the Trilogy (p=0.04). The difference between the HG1 and HG2 was not significant. Despite the limitations imposed by the small sample studied, the presence of multiple screw holes in the Harris-Galante retrievals, and sub-optimal matching for sex, height, and varied indication for revision, we detected significant reduction in the backside wear of modern modular and non-modular acetabular cups when compared to first generation modular designs.

Particles generated at the non articulating surface (backside) of modular acetabular components have been implicated in the development of periprosthetic osteolysis after THA. Several design changes have been introduced in modern acetabular cups in an attempt to reduce backside wear, including the use of “non-modular cups”. We compared the backside wear of retrieved cementless non-modular cups, with modular cups of first and second generation designs. Nine retrieved non-modular cups (Implex) were match-paired for time-in-situ, patient age and weight, with 9 retrieved Trilogy cups, 9 Harris-Galante 1, and 9 Harris-Galante 2. The average time in situ was 2.5 years (1–7). The backside was divided in quadrants and each rated with a value from 0 (absence of wear) to 3 (severe backside wear) for a total ranging from 0 to 12. This new score was validated for intra and inter observer reproducibility. Among 36 quadrants in the HG1 group there were 3 rated 1, 23 rated 2, and 10 rated 3. In the HG2 group, 1 quadrant was rated 0, 16 rated 1, 14 rated 2, and 5 rated 3. In the Trilogy group, 6 quadrants were rated 0, 27 rated 1, and 3 rated 2. In the Implex group, 15 quadrants were rated 0, 21 rated 1. The average backside wear score and 95%CI were 8.4 (7.6–9.3); 7.3 (5.5–9.1); 3.7 (3.2–4.1); and 2.3 (1.3–3.4) respectively. The HG cups demonstrated more severe backside wear than the Trilogy and Implex (p< 0.02). There was a tendency towards less backside wear in the Implex cup when compared to the Trilogy (p=0.04). The difference between the HG1 and HG2 was not significant. We detected significant reduction in the backside wear of modern modular and non-modular acetabular cups when compared to first generation modular designs.

Bankston et. al. reported that the clinical wear rates of molded acetabular cups was 50% less than a group of machined UHMWPE cups. However, due to covariables between groups including different femoral stems, cement technique, polyethylene resins and surgeons, unequivocal attribution of the low wear rates to direct molding could not be made.

In order to more directly assess the benefits of directly molded acetabular cups vs. machined cups, we report the comparison of hip simulation wear rates of machined and directly molded cups. These simulator results will then compared to two recent clinical reports on molded and machined cups of the same hip stem and cup design. The molded cups were made from 1900 resin and gamma sterilized in an inert atmosphere. The machined cups were made from HSS reference UHMWPE (4150) and gamma sterilized in air. The molded 1900 cups had a 55% lower wear rate after 5 million cycles on the hip simulator (14 v. 31mg/million cycles). Ranawat reported the average linear head penetration rate for 235 direct-lymolded, all polyethylene, cemented cups at a mean follow-up time of 6 years was .075mm/year. This is 56% lower than the rate of .17 mm/year he reported previously for the machined, uncemented metal-backed cups of the same design. These results provide further evidence that directly molding acetabular cups can provide wear rates over 50% less than machined cups both in both clinical and hip simulator evaluations. It is interesting to note based on other reports, that there is no osteolysis at 10 years of follow-up when the wear rates are < 1mm. The clinical and simulation wear rates reported here for the directly molded cups are within this performance range.

Improving the wear resistance of polyethylene is considered paramount to improving knee implant longevity. Consequently, a range of polymer fabrication techniques have evolved in the quest for a highly wear resistant material. The objective of this study was to explore the wear performance of polyethylene as fabricated in a variety of ways.

The following materials were prepared, sterilised, artificially aged, and machined into wear specimens (n = 4 for each material): Compression molded GUR1050 with three levels of cross-linking (120 kGy, 65 kGy, and 0 kGy irradiation – control); ram extruded GUR4150 high modulus material; compression molded GUR4150 low modulus material; and HSS/PolySolidur/Hoechst reference polyethylene. Using a custom designed joint articular wear simulator, samples were loaded for 2 million cycles at a frequency of 0.5 Hz under loads of 2.1 kN. Tests were stopped every 250 000 cycles; and wear surfaces were examined microscopically for surface damage (pitting, cracking, delamination).

After 2 million loading cycles the following specimens were pitted and delaminated: 2 GUR1050 control samples, 3 GUR4150 high modulus specimens, and all 4 reference polyethylene specimens. Burnishing, but no pitting, was seen in all GUR1050 elevated cross-linked polyethylene specimens, and in all GUR4150 low modulus specimens.

The materials tested in this study represent a broad range of fabrication techniques. Differences in starting resin cannot fully account for the differences in wear behaviour seen between the groups; as damage was not limited to one resin group. The cross-linked specimens were melt-annealed, prior to cross-linking. It is possible that this processing step, and not the actual cross-linking, contributed to the improved wear performance of this group. However, of most interest is the comparable wear performance of GUR1050 cross-linked polyethylene and GUR4150 low modulus polyethylene suggesting that cross-linking polyethylene is not the only route towards obtaining a polyethylene with superior wear characteristics.