Résumé
Background: Posterior humeral subluxation is the main cause of failure of total shoulder arthroplasty. We aimed to compare humeral head subluxation in various reference planes and to search for a correlation with retroversion, inclination, and glenoid wear.
Materials and methods: We included 109 computed tomography scans of primary glenohumeral osteoarthritis and 97 of shoulder problems unrelated to shoulder osteoarthritis (controls); all computed tomography scans were reconstructed in the anatomic scapular plane and the glenoid hull plane that we defined. In both planes, we measured retroversion, inclination, glenohumeral offset (Walch index), and scapulohumeral offset.
Results: Retroversion in the scapular plane (Friedman method) was lower than that in the glenoid hull plane for controls and for arthritic shoulders. The threshold of scapulohumeral subluxation was 60% and 65% in the scapular plane and glenoid hull plane, respectively. The mean upward inclination was lower in the scapular plane (Churchill method) than in the glenoid hull plane (Maurer method). In the glenoid hull plane, 35% of type A2 glenoids showed glenohumeral offset greater than 75%, with mean retroversion of 25.6 degrees +/- 6 degrees as compared with 7.5 degrees +/- 7.2 degrees for the "centered" typeA2 glenoids (P < .0001) and an upward inclination of -1.4 degrees +/- 8 degrees and 6.3 degrees +/- 7 degrees (P = .03), respectively. The correlation between retroversion and scapulohumeral offset was r = 0.64 in the glenoid hull plane and r = 0.59 in the scapular plane (P < .05).
Conclusion: Measurement in the glenoid hull plane may be more accurate than in the scapular plane. Thus, the glenoid hull method allows for better understanding type B3 of the modified Walch classification. (C) 2017 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved.