Purpose: This study was conducted on explanted uni-compartmental prostheses with a flat polyethylene plateau without metal backing. We search for clinical factors influencing polyethylene wear.

Material and methods: This series included 30 polyethylene inserts divided into two groups. Group A included revision procedures performed for a reason other than implant loosening (wear of the other compartment, femoropatellar problems). Group B implants were explanted after loosening. The duration of implantation of the 13 implants in group A was 126 months (mean, range 11–218 months); it was 167 months (range 137–224) in group B. Remaining insert thickness was measured with a micormetric device, mitutoyo, allowing palpation of the worn surface with a precision of 3 microns. The volume of the femoral penetration into the polyethylene was calculed in two ways to separate penetration related to polyethylene deformation from penetration related to polyethylene wear. The micrometric device palpated the surface of the polyethylene enabling calculation of the sum of the volumes corresponding to wear and deformation. To measure the volume corresponding to wear, the explanted pieces were weighed and the result was compared with implants of the same size which had never been implanted. Polyethylene wear was calculated from the weight loss and and polyethylene density. The difference between the two calculation methods was attributed to polyethylene deformation.

Results: Mean residual thickness of the polyethylene in group A without loosening was 7.16 mm, compared with 4.5 mm in group B. The volumetric femoral penetration into the polyethylene was a mean 19 mm3 per year in group A and 65 mm3 per year in group B. This imprint obtained with the micormetric measuring device was greater than the wear determined by weighing. This difference was about 25%. The decreased thickness of the implant was thus undoubtedly due, for three quarters, to wear alone. One quarter being attributed to polyethylene deformation. In group A (without loosening), each supplementary year of implantation corresponded to a decrease in the rate of wear of about 12% per yar, which would suggest that the wear mechanism is an abrasion and that with time the femoral and tibial implants become more congruent decreasing the rate of wear. Inversely in group B, each supplementary year of implantation after onset of loosening was associated with a 9% increase in the annual rate of wear. Microscopic examination of the group A implants demonstrate that abraison was the main mechanism of wear. In group B, delamination was observed, particularly when the loosening was associated with anterior cruciate ligament tear or major persistent deformation.

Discussion: The rates and mechanisms of polyethylene wear in unicompartmental prostheses are different for non-loosened implants and for loosened implants. Taking into account the fact that polyethylene deformation participates for about one quarter of the decreased thickness over time, adjunction of a metal back would appear indispensable for thin inserts.