Comparison of in vitro with in vivo characteristics of wear particles from metal–metal hip implants

The purpose of the present study was to compare wear particles isolated from metal–metal (MM) hip implants worn in an orbital bearing simulator with particles from similar MM total hip replacement (THR) implants worn in vivo . Comparison of these particles is important because it will help to assess the overall suitability of this type of hip simulator for reproducing in vivo wear and for producing physiological wear particles suitable for biological studies of in vitro cellular response. Commercial grade components made of ASTM F75 (cast) alloy were evaluated. Simulator tests were performed in 95% bovine calf serum with a 28‐mm‐diameter implant. Wear particles were collected from 0 to 0.25 million cycles (run‐in wear period) and 1.75 to 2 million cycles (steady‐state wear period). Tissues from seven patients with MM implants (surface replacement or stem type) were harvested at revision surgeries (after 1–43 months). Metal wear particles were isolated from serum lubricant or tissues using an enzymatic protocol that was previously optimized to minimize particle changes due to reagents. After isolation, particles were centrifuged, embedded in epoxy resin, and characterized by transmission electron microscopy (TEM) and energy dispersive X‐ray analysis (EDXA). Results of EDXA on particles from the hip simulator primarily indicated a predominance of particles containing Cr and O but no Co (most likely chromium oxide particles), and fewer CoCrMo particles presenting varying ratios of Co and Cr. Image analysis of TEM micrographs demonstrated that the majority of the particles from the simulator were round to oval, but a substantial number of needle‐shaped particles were also found, especially from 0 to 0.25 Mc. The particles generated from 0 to 0.25 Mc had an average length of 53 nm, whereas those generated from 1.75 to 2 Mc had an average length of 43 nm. In vivo , EDXA and TEM analysis of particles that were retrieved from two patients at 23 and 43 months respectively, revealed that they were the most comparable in composition, average length (57 nm), and shape to particles generated in the hip simulator during the run‐in wear period. Because a large clinical retrieval study in the literature suggested that a run‐in wear regime might occur in vivo for some 6–36 months, the fidelity of the simulator of the present study was strongly supported. However, some uncertainties existed, including the finding that the particles isolated from the other five patients generated from 1 month up to 15 months (shorter implantation times than the other two patients) were smaller and mostly contained only Cr and O (no Co). In the opinion of the authors, this particular very short term patient group was somewhat atypical. Therefore, despite these uncertainties, the present study was deemed to support the ability of the orbital bearing hip simulator to produce physiological wear particles. © 2004 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater 70B: 167–178, 2004