BACKGROUND

Our purposes were to analyze radiographic changes around humeral components and to determine the influence of these changes on the clinical outcome of shoulder arthroplasty.

METHODS

Three hundred and ninety-five shoulders with primary osteoarthritis were treated with a shoulder replacement and were analyzed clinically and radiographically with a standardized protocol. Radiographs were evaluated for osteolysis and stress-shielding of the proximal part of the humerus and loosening of the humeral and glenoid components. The arthroplasty configurations included (1) hemiarthroplasty with cement (thirty shoulders), (2) hemiarthroplasty without cement (thirty-six), (3) total shoulder replacement with a cemented stem and a cemented flat-backed glenoid component (eighty-nine), (4) total shoulder replacement with a cemented stem and a cemented convex-backed glenoid component (ninety-four), (5) total shoulder replacement with a cemented stem and an uncemented metal-backed glenoid component (seventy-nine), and (6) total shoulder replacement with an uncemented stem and a cemented convex-backed glenoid component (sixty-seven). The mean duration of follow-up was 8.2 years (range, four to eighteen years).

RESULTS

The Constant score and shoulder motion were significantly greater at the latest follow-up examination than preoperatively (p < 0.001). One cemented stem loosened after a periprosthetic fracture, and another was judged to be at risk for loosening. Stress-shielding was noted only around uncemented stems in 63% of the shoulders with such stems. Osteolysis of the proximal part of the humerus occurred only with total shoulder arthroplasty in 43% of the shoulders. Patients with osteolysis had a poorer clinical outcome (p < 0.001). Shoulders with osteolysis around the proximal part of the humerus had more glenoid loosening and polyethylene wear (p < 0.001).

CONCLUSIONS

Glenoid issues have a substantial effect on humeral bone in shoulder arthroplasty. Polyethylene wear appears to be associated with the development of osteolysis of the proximal part of the humerus after total shoulder arthroplasty. With our implants, stress-shielding was only observed with uncemented humeral stems. The uncemented and cemented stem designs analyzed in this study appeared comparable and can be recommended for clinical use.