Shoulder hemi arthroplasty radiological and clinical outcomes at more than two years follow-up

SummaryIntroductionHumeral head replacement is used for glenohumeral osteoarthritis in young or active patients, for conditions sparing glenoid cartilage or when glenoid implantation does not appear feasible. These surgical procedures usually give satisfactory results but there is a risk of glenoid erosion and a possible deterioration of long-term outcomes.HypothesisThere is a risk of glenoid erosion after humeral head replacement which can be radiographically measured. The importance and progression of this erosion should be evaluated to determine its clinical relevance.Patient and methodsThis is a retrospective study in 15 patients (19 shoulders) who underwent humeral head replacement between 1999 and 2006. There were 11 women and four men with an average age of 54.5 years. Etiologies were avascular necrosis (11 cases) and glenohumeral osteoarthritis (eight cases). All patients were reviewed in 2008 with more than two years follow-up. Clinical evaluation included measurements of range of motion and determination of the Constant and Murley score. In addition, the patients were asked to provide a subjective evaluation of their shoulder. Radiographic analysis included computer-assisted measurements.ResultsThe average follow-up was 45.8 months (26–108). At one year postoperative and at the final follow-up, clinical parameters such as the Constant and Murley score (37.4/100 preoperative to 64.4/100 at final follow-up) were significantly increased. During the first year, the rate of glenoid wear was 1.03 mm/year in case of avascular necrosis and 0.27 mm/year in case of osteoarthritis (p < 0.001). Glenoid depth at the final follow-up was 6.97 mm for osteoarthritis compared to 4.59 mm for avascular necrosis (p < 0.01). We did not find any correlation between glenoid erosion severity and clinical results.DiscussionIsolated humeral head replacement may result in glenoid erosion. The rate of progression of this erosion is clearly influenced by the etiology and therefore by the preexisting condition of the glenoid cartilage. At the average follow-up, the radiological glenoid deterioration is not correlated with pain or deterioration of clinical results.Level of evidenceLevel IV. Therapeutic study.