The Appropriate Variable to Compare Wear in Total Hip Prostheses

Commentary An important limiting factor in the longevity of any total joint replacement is its propensity to wear. Unlike bearings such as ball bearings and hydrodynamically lubricated bearings that produce virtually no wear, sliding bearings that have contact surfaces with relative motion will always wear. For total hip replacement prostheses, wear occurs with all of the currently used technologies. We study wear in these prostheses because of the 2 failure modes associated with it: loss of mechanical integrity and, more commonly, wear-induced biological failure. The latter failure mode is the most important clinical problem facing the bearing designer, the treating surgeon, and the patient. It is, therefore, important that the models used to predict the longevity of a particular bearing design are based on the most logical scientific evidence available. The most valuable evidence for predicting longevity is based on clinical observation. Toward this end, the observations presented by Howie et al. are extremely valuable not only for the conclusions derived from the study but also for the wear data itself. When using wear data to derive clinical prediction models, it is important to keep in mind that the more similarity between prostheses, the more simple and basic can be the wear parameter. Thus, for similar materials and identical bearing diameters, a comparison of linear wear as measured by radiographic penetration is useful and proper. If we change the diameter of the bearings, then linear wear is not an appropriate comparison variable; rather, the radiographic measurement must be converted into volumetric wear. For conventional polyethylene bearings, this concept is well described and validated by, for example, Oparaugo et al.1. Thus, when Howie et al. state that the “wear rate” is not increased by large articulations, they are in fact referring to the linear wear rate or penetration rate. When they calculate the volumetric wear rates, there is an approximate 2:1 ratio between 28-mm articulations and 36-mm articulations. If the reader is to use the information presented in this article to estimate the likelihood of the future occurrence of osteolysis for a metal-on-polyethylene bearing articulation, then the volumetric wear rate should be used. Indeed, the literature supports the hypothesis that volumetric wear rates from conventional polyethylene bearings may be predictive when applied to the slightly modified materials in highly cross-linked polyethylene bearings. Thus, the volumetric wear rate will suffice as an acceptable measurement variable when evaluating the potential for osteolysis development with metal-on-polyethylene bearings. The same cannot logically be held to be true if we are to model metal-on-metal bearings. That is, volumetric wear rates based on polyethylene bearings may have no correlation to volumetric wear rates of metal-on-metal bearings with regard to the biological viability of the joint. The biological basis for this statement is that osteolysis and related pathologic tissue transformations are particle-induced diseases. It is well recognized that in classic tissue-culture models, in which cells are challenged by particles, the number of particles is the important variable rather than the volume of material. As a result, when we cross into the realm of metal particle-induced bioreactivity, we must recognize that particle size and composition influence the outcome. Thus, while volumetric wear of a metal-on-metal bearing may be considerably lower than that of a metal-on-polyethylene bearing, direct comparison of this variable is meaningless because metal particle debris is much smaller than polyethylene particle debris. Consequently, smaller volumes of metal can produce numbers of particles that are orders of magnitude larger than those produced by polyethylene bearings. In conclusion, wear rates that are expressed as the linear wear rate or penetration rate are useful as a direct comparison only if both the material and diameter of both populations are the same. For different diameters of polyethylene bearings, the volumetric rate should be the variable being considered. When comparisons are made across bearing materials, such as between metal-on-polyethylene and metal-on-metal bearings, volumetric wear rates are no longer useful as a comparator, since equal volumes result in orders of magnitude more metal particles. That the bioreactivity of metal particles differs markedly from that of polyethylene particles further precludes the use of direct volumetric wear rate comparisons.