Introduction
Loosening of the glenoid component after a total shoulder arthroplasty (TSA), leading to pain and the need for a revision operation, accounts for one third of all shoulder replacement complications. It has been shown that glenoid loosening could be a result of osteolysis of periprosthetic bone induced by polymeric wear debris.
The aim of this study is to determine the amount and pattern of in vivo wear of polyethylene glenoid components by examining a set of so called retrievals using laser scanner technique and scanning electron microscopy.

Methods
Six glenoid components were retrieved and examined (five are all-polyethylene and one metal-backed). The average time of functioning of these components in human bodies was 10 years with the range of values from 7-14 years.
Each component was scanned using a laser scanner. The results of measurements have been used to generate geometrical models of the retrievals using CAD software - GEOMAGIC. These models have been subsequently used to determine the maximum depth and the volumetric of wear for each component. The retrievals were also examined using the scanning electron microscope (SEM).

Results
The maximum wear rate and wear depth for the analyzed retrievals was 30 mm3/yr and 0.2 mm/yr, respectively. From the SEM analysis of the retrievals it was found that prominent inferior wear areas and fine multidirectional scratches were dominant in all retrievals. Random scratches were also observed, most probably due to the third body wear (cement particles). Two implants were found to have surface micro-cracks, believed to result from subsurface fatigue.

Discussion
The observed wear rates and depths in the glenoid prosthesis are within the average value observed clinically in hip arthroplasty. The significant abrasive and fatigue wear observed in the study might be the result of high surface and and subsurface stresses acting in TSA and multidirectional component relative movement.