The objective of this study was to compare the wear mode of 100 Mrad PE cups run in a hip simulator to retrieved 100 Mrad PE cups, and to evaluate the efficacy of the PE wear model.

15 In-vitro PE cups: 3 each 0,2.5.50,100 and 150 Mrad (9 channel hip simulator, 6.2 million cycle duration, physiological load profile by Paul, 2000N maximum load at 1Hz using 30% bovine serum). 5 Retrieved PE cups: three SOM cups (Mizuho Medical Instrument Co., COP alloy 28 mm head)-0 Mrad after 8 years of clinical use, two 100 Mrad cups after 15 years of clinical use, two T28 PE 2.5 Mrad cups (Zimmer): 18 years and 13 years of clinical use. The cups were examined using a SEM (Philip XL30 FEG) for wear scar locations and PE wear-topography.

Original machine marks were observed in the weight-bearing areas of the highly cross-linked in-vitro PE. No machine marks were observed for the 0 and 2.5 Mrad in-vitro cups and none were seen in any of the retrieved cups. The formation of more nodules and fibrils in the 0Mrad cups compared to the extensivley cross-linked cups (in-vitro and retrieved) was striking. The frequency of occurrence and length of the fibrils and nodules was dependent on the dose of gamma irradiation. More ripples were formed in the 2.5 Mrad and higher cups compared to the non-irradiated cups (in-vitro and retrieved). The in-vitro cups formed more ripples than the retrieved cups. In general, the SEM features for in-vitro Mrad cups appeared similar to those of the retrieved Mrad cups.

The in-vitro Mrad cups accurately reflected the conditions of the artificial joint in living body. Therefore, comparisons of retrieved PE cups with simulator PE cups appeared to be a very powerful research tool.

(2) SEM observation demonstrated far less wear damage in the extensively cross-linked cups than in the non-extensively cross-linked PE. Thus the extensive cross-linked PE cups appeared to be a significant improvement over conventional PE cups in terms of wear resistance.