Primary Gleno-humeral Osteoarthritis: Clinical and Radiographic Classification

Abstract

One hundred and fifty-one cases of primary glenohumeral osteoarthritis were analysed clinically and radiographically with computed tomography. There were 76% women and the mean age at operation was 66.8 years. Diagnosis was made only with true anteroposterior views of the shoulder which show the narrowing of the gleno-humeral joint and the presence of osteophytes. Arthrography showed a supraspinatus tear in 10% of the cases which extended to the infraspinatus in 2.6%. Glenoid retroversion averaged 15.4°. However, measurements of glenoid retroversion with computed tomography do not allow posterior wear to be distinguished from glenoid dysplasia. Posterior subluxation of the humeral head was observed in 45% of the cases. We describe a classification of the glenoid morphology in three types: Type A (53.5%) is characterised by an equal balance of forces acting on the glenoid and a centralised humeral head. Type B (39.5%) has asymmetrical posterior force distribution on the glenoid and a posterior subluxation of the humeral head. Type C (5%) is arbitrarily defined as glenoid retroversion greater than 25°. Posterior subluxation of the proximal humeral head was not correlated with either glenoid retroversion or humeral retroversion. However, subluxation of the humeral head may be responsible for the posterior glenoid erosion and the possible biconcave appearance of the glenoid observed in primary gleno-humeral osteoarthritis (Type B). The posterior subluxation of the humeral head may still be present after shoulder arthroplasty and may be responsible for glenoid loosening due to a “horizontal rocking horse effect”.