Clinical and radiographic long-term outcomes of hemiarthroplasty for complex proximal humeral fractures

Abstract

Hemiarthroplasty (HA) is a treatment option for complex proximal humerus fractures not suitable for conservative treatment or open reduction and internal fixation (ORIF). Long-term outcomes using a large metaphyseal volume prosthesis in the management of fractures of the proximal humerus have not been reported thus far. Between 2006 and 2010, 41 patients with a proximal humerus fracture were treated with HA (average age: 62 years: range, 38-85) at our institution. Nine patients underwent revision surgery, three patients were lost to follow-up and seven died unrelated to the index surgery. Twenty-two patients were reviewed clinically and radiographically at a mean period of 10.4 years (range, 9-13 years). Seven out of nine failures of HA occurred within the first 2 postoperative years: 2 infections and 5 greater tuberosity non-/mal-unions. The other two patients underwent revision for rotator cuff deficiency more than 5 years after initial surgery. Of those patients available for final follow-up, the implant survival rate was 71% (22 out of 31 patients). Their mean relative Constant score was 76 % (range, 49-96), the mean active elevation was 116° (range, 60-170) and the mean external rotation was 28° (range, 0-55), at the final follow-up. The majority had a good/excellent internal rotation, with 13 patients (59%) to the 12th thoracic vertebra and 7 patients (31%) to the 8th thoracic vertebra. The mean Subjective Shoulder Value was 76% (range, 40-100). Clinical outcomes did not significantly deteriorate over a period of 10 years, except for flexion (p<0.001) and internal rotation (p=0.002). Analyzing greater tuberosity healing, one patient had a non-union, 10 (45%) had a mal-union, and in 12 patients (55%) the greater tuberosity healed in an anatomical position. Patients with a displaced mal-union of the greater tuberosity did not have inferior clinical results at last follow-up. Only two patients showed glenoid erosion and in no patient stem loosening could be identified at the final follow-up. The revision rate following large metaphyseal volume HA to treat a proximal humerus fracture was 29% after 10 years postoperatively, with failure within 2 years largely related to greater tuberosity non-/mal-union and later related to rotator cuff insufficiency. Patients with a retained implant showed good clinical and radiographic long-term results, without relevant deterioration over time even when the greater tuberosity healed in a non-anatomical position.