THE FALLACY OF EVALUATING BIOMATERIAL WEAR-RATES WITH WATER AS LUBRICANT - A HIP SIMULATOR STUDY OF ALUMINA-PTFE AND COCR-PTFE COMBINATIONS

Controversy surrounds wear data from hip-simulator studies, whether fr om the choice of lubricants or other parameters such as the particular biomaterial combinations used, and whether any such interactions coul d bias the resulting wear predictions. To investigate these phenomena, we studied the wear performance of CoCr and alumina femoral heads, in water and serum-based lubricants, using as our standard the polytetra fluoroethylene wear data derived clinically by Charnley. To model Cham ley's clinical experience, PTFE acetabular cups were used in sets of t hree each with each size of femoral head for 22.25, 28, and 42-mm diam eters in a nine-channel hip simulator. From the serum-based tests, the CoCr-PTFE wear data were consistently linear with duration of test, e xhibited very large wear rates of 3,000-8,400 mm(3)/10(6), cycles had a precision within +/- 4% for each set of three cups, and copious amou nts of small particulate were clearly seen circulating. The wear data clearly demonstrated Charnley's thesis that volume of wear increased w ith regard to size of femoral head. From the waterbased tests, the CoC r-PTFE wear data were nonlinear with duration of test, had much reduce d wear rates compared to the serum tests, lost the clinical relationsh ip with ball size, and precision deteriorated to +/- 27% for each set. The wear debris appeared as 1-2 cm long ribbons which floated to the surface. For the alumina-PTFE combination in serum, the wear data appe ared identical in performance to the CoCr-PTFE data in serum. Thus, th e PTFE wear rates were not sensitive to the choice of femoral-head mat erial. The most surprising outcome in this study was the zero-wear per formance of the ceramic-PTFE combination in water. This contrasted rem arkably with the large wear rates established for the same combination s run in serum. The zero-wear performance of the ceramic-PTFE combinat ion in water was unexpected, but a similar phenomenon was noted in pub lished simulator tests of ceramic-UHMWPE run in water. It now seems Li kely that such data may reflect the capricious behavior of water lubri cation rather than any other variables under evaluation. The water-bas ed experiments clearly favored the ceramic's superior tribological per formance and placed metal bearings at a decided disadvantage. Therefor e, for an in vitro simulation of materials wear-ranking of clinical re levance, it may be advisable to use a serum-based lubricant. (C) 1998 John Wiley & Sons, Inc.