Arthroplasty for glenohumeral arthropathies have specific complications and the final results are sometimes more dependent upon the type of shoulder arthroplasty than the initial etiology. The aim of our study was to evaluate the rate of complications and the functional improvement with different types of shoulder arthroplasties after a minimum follow-up of 8years.

This was a multicenter retrospective study of 198 shoulders including 85 primary osteoarthritis of the shoulder, 76 cuff tear arthropathies, 19 avascular necrosis and 18 rheumatoid arthritis. Arthroplasties included 104 anatomic total shoulder arthroplasties (TSA), 77 reverse arthroplasties and 17 hemiarthroplasties. Ten patients had their arthroplasty revised, and 134 patients with TSA were able to be present at the final follow-up or provide information on their case. Function was evaluated by the Constant-Murley score and loosening by standard radiographs.

In the group with primary osteoarthritis of the shoulder, there were eight complications (11%) including six (8.3%) requiring implant revision. In the group of rotator cuff arthropathies, there were nine (14.7%) complications including four (6.5%) requiring implant revision. In the group with rheumatoid arthritis, there was one complication, and no surgical revision was necessary. There were no complications in the group with avascular necrosis. Glenoid migration occurred in 28.5% of anatomic TSA, and 3.4% of reverse arthroplasties. This difference was significant (P<0.001). The Constant-Murley score was significantly improved in all etiologies.

Glenohumeral arthropathies can be successfully treated by arthroplasty. Anatomic TSA was shown to be associated with a high risk of glenoid loosening at radiographic follow-up, which makes us hesitate to use the cemented polyethylene implant, especially in young patients.

IV – Retrospective study.