Wear of Co-Cr-Mo interfaces for 1-20 year THA retrievals

The generation of particulate debris from TJA prosthetic surfaces has been shown to incite biological responses and result in loosening of prosthetic components. Thus, there is renewed interest in metal-metal articulations as a method of minimizing wear debris and improving the longevity of the implant. The purpose of this study was to quantitate the wear over time of Sivash/SRN total hip arthroplasties (THA). Twenty-six retrieved and three non-implanted Sivash/SRN Co-Cr-Mo articulating surface THA devices were evaluated in this study. Among the 29 devices, 21 femoral heads, 28 acetabular cups, and 20 headcup matches were analyzed. The average time of implantation was 7.2 years, with a minimum time of 1.33 years and a maximum of 20. Wear was quantified by a three-dimensional surface analysis technique which utilized a Checkmaster coordinate measuring machine and Geomet software. The nominal radius and sigma (/spl sigma/) values were calculated for each cup and head. The error between each surface data point and the nominal sphere was also calculated. Deviations from the nominal sphere revealed areas of surface irregularities due to machining and/or wear, and wear volumes were determined by multiplying regional wear depths by their corresponding areas. A correlation analysis and a general linear model procedure were performed on each of the following variables versus time: head radius, head (/spl sigma/ value, cup radius, cup /spl sigma/ value, and cuphead fit (radius of cup minus radius of head). Because wear in THA has been found to be polar, significant wear should produce larger deviations from the original nominal sphere, i.e. larger /spl sigma/ values. Therefore, the annual volumetric wear rates of three cups and three heads, each demonstrating the largest /spl sigma/ value in their respective age category, were calculated. Correlation analyses revealed low correlation coefficients between time of implantation and all of the variables tested. No statistically significant trends were observed (p<0.05) between time of implantation and any of the variables tested using a general linear model procedure.