Abstract
Materials and methods: From 1993 to 2001 in 20 patients with rheumatoid arthritis 24 total shoulder arthroplasties were performed. Concerning 19 Biomet total shoulder arthroplasties of which one glenoid component and one humeral component were cemented for optimal fixation. Five Aequalis total shoulder arthroplasties were performed, one with an uncemented glenoid component. The pre-operative diagnosis was rheumatoid arthritis in the whole Biomet-group. In the Aequalis-group there were four patients with rheumatoid arthritis and one patient with juvenile chronic arthritis. The mean age was 48 (19–76) years. The mean pre-operative range of motion was flexion 80 (40–150), external rotation 23 (–20 – 65) and abduction 55 (0–110) degrees. Pre-operative shoulder X-rays were staged according to Larssen. In the Biomet-group this resulted in eight times Larssen 3, nine times Larssen 4 and two times Larssen 5; in the Aequalis-group five times Larssen 4.
All 24 operations were performed by a deltopectoral approach with 11 longitudinal osteotomies of the clavicula, 12 osteotomies of the lesser tubercle, two osteotomies of the greater tubercle and two osteotomies of the coracoid process; all without complications. Seven times a cancellous bone graft of the glenoid was performed. A rotatorcuff rupture was seen five times; in all cases this was closed primarily. A rotatorcuff release was performed four times. Two procedures were complicated: one longitudinal fissure of the humerus and one fractured greater tubercle, both in the Biomet-group. The after-treatment for all shoulder arthroplasties was functional with use of a collar ‘n cuff accompanied by intensive physiotherapy with increasing load of use.
Results: No (sub)dislocation was seen or revision surgery was performed within six weeks after operation. Eight cranial subdislocations and one caudal subdislocation occured and one lesion of the brachial plexus was diagnosed six weeks or more after operation in the Biomet-group. There were no complications in the Aequalis-group. The mean follow-up is 49 (4–93) months. The mean postoperative range of motion was flexion 79 (20–150), external rotation 36 (0–75) and abduction 62 (0–150) degrees. The mean visual analogue painscore (VAS) increased from 3, 3 pre-operatively to 7, 3 postoperatively. Radiolucent lines were seen nine times, cemented Biomet humeral component 1x, uncemented Biomet glenoid component 7x and around cemented Aequalis glenoid component 1x. Two uncemented Biomet glenoid components showed medial migration. None humeral components showed migration. Two times a screw breakage was seen. Revision surgery was performed in two patients with subdislocation in the Biomet-group; a thicker humeral head was placed both times. One with good result and one re-subdislocated due to a lesion of the brachial plexus.
Discussion: in 24 total shoulder arthroplasties no loosening of the humeral component occured. Loosening was seen in two glenoid components (8, 3%) although there was no indication to perform revision surgery. In these two cases a decreasing shoulder function was found. According to the pre-operative situation the range of motion post-operatively was unchanged but pain was obviously less.
