Dimensional characteristics of uncomplicated autopsy-retrieved acetabular polyethylene liners by ultrasound

A new, in vitro ultrasound-based method to measure the thickness of acetabular polyethylene components was developed and applied to 26 uncomplicated autopsy retrieved components and 40 unused components. The average age at total hip arthroplasty was 62 years and the average time in service of the retrieved components was 49 months. The clinical notes indicated that each of the arthroplasties was functioning well at the time of the patients death. Thickness distributions in the retrieved components had two distinct patterns. Eighteen of the retrieved components (69%) had their thinnest areas self-contained and located near the polar point. In the other 8 retrieved components (31%) the areas of minimum thickness were adjacent to the rim. Thickness distribution in the unused cups was predominately concentric with the thinnest area located near the polar point (85% of the cups). Detection limits for dimensional change were established based on the variability found in the unused liners. Fifteen of the 26 retrieved components (58%) had areas of reduced thickness which exceeded the detection limits; the average thickness reduction rate for these components was 0.076 mm per year. The other 11 retrieved components (42%) lacked such areas. The 15 cups with areas of reduced thickness had a longer time in service (63 +/- 18 months) than the 11 cups without areas of reduced thickness (32 +/- 25 months) (p = 0.003), but no other clinical factor (age, gender, Harris hip score, size and inclination of the cup, type of femoral fixation) was associated with these 15 cups. Cylindrical models to estimate volumetric change tended to underestimate the actual changes, suggesting that the actual particulate burden may be greater than previously appreciated. Finding that the pattern of thickness reduction can vary suggests that distinctive hip loading modes may be present postoperatively in patients with total hip arthroplasty. The wear rates of these components are consistent with wear rates calculated from radiographic data for well-functioning implants and are considerably lower than wear rates calculated for surgically-retrieved implants, indicating that autopsy-derived retrievals may be more representative of the majority of components currently in service than surgically-derived retrievals.