Radiographic assessment of cemented humeral components in shoulder arthroplasty.

Forty-three shoulder arthroplasties performed with the use of cemented Neer II humeral components and followed radiographically for a mean of 6.6 years (range, 2-20 years) were analyzed. A humeral component was considered radiographically "at risk" for clinical loosening when a radiolucent line 2 mm or greater in width was present in 3 or more zones or tilt or subsidence was identified on sequential radiographs by 2 or 3 of the 3 independent observers. None of the components was considered to have tilted or subsided. Radiolucent lines of any size were present in 16 shoulders and were wider than 2 mm in 9 shoulders. They were limited to 1 zone in 8 shoulders and to 2 zones in 7 shoulders. Only 1 component (2%) with a 2-mm radiolucent line in 3 zones was judged to be "at risk." The incidence, extent, and thickness of humeral radiolucent lines were significantly higher in total arthroplasties than in hemiarthroplasties (P <.05). Clinically important changes around cemented Neer II humeral components are uncommon. Humeral radiolucent lines develop more frequently in the presence of a glenoid component. Data from this study can be used as one benchmark to compare with alternate methods of humeral component fixation.