Risks of loosening of a prosthetic glenoid implanted in retroversion Alain Farron, MD, a Alexandre Terrier, PhD, b and Philippe Büchler, PhD, b Lausanne, Switzerland Osteoarthritis of the shoulder is frequently associated with posterior glenoid wear, which may be difficult to correct during shoulder arthroplasty. This study was designed to evaluate the risks that a prosthetic glenoid implanted in retroversion will loosen. The scapula, the humerus, the rotator cuff, and a total shoulder prosthe- sis were reconstructed with a 3-dimensional finite ele- ment model. The glenoid was placed in 5 different angles of retroversion (0°, 5°, 10°, 15°, and 20°). Location of the glenohumeral contact point, articular pressure, bone and cement stress, and micromotion around the glenoid implant were calculated during internal and external rotation. Glenoid retroversion induced a posterior displacement of the glenohumeral contact point during internal and external rotation, in- ducing a significant increase of stress within the ce- ment mantel (+326%) and within the glenoid bone (+162%). Furthermore, a major increase of micromo- tion was measured at the bone-cement interface (+706%). According to this study, glenoid retroversion exceeding 10° should be corrected during total shoul- der arthroplasty. If the correction is impossible, not replacing the glenoid should be considered. (J Shoul- der Elbow Surg 2006;15:521-526.) O steoarthritis of the shoulder is frequently associated with posterior glenoid wear. 6,15,23 The reasons remain unclear, but are probably multifactorial. The excessive stiffness of the anterior soft tissues after previous surgical procedures performed through an anterior approach is classically recognized as a cause of posterior glenoid erosion. 1,7,13 Posterior glenoid wear is also frequently seen without any previous surgery, however. In recent years, primary glenoid dysplasia, associated with in- creased glenoid retroversion and sometimes with a static posterior subluxation, has also been evoked as a cause of shoulder osteoarthritis. 24,25 Among the other possible causes, the effects of muscular imbalance (eg, in association with neurologic problems) and the influ- ence of the humeral side of the joint (especially retrover- sion of the humeral head) are less frequently reported. During shoulder arthroplasty, correction of a retro- verted or posteriorly eroded glenoid is a difficult task. Reaming of the anterior part of the glenoid 14 may significantly reduce bone stock, leading to concerns about the stability of the prosthetic component. Fur- thermore, this procedure will medially displace the center of rotation, which may provoke impingement between the coracoid process and the humeral head. Bone grafting of the posterior glenoid, 9,16,19 associ- ated or not with an osteotomy, is an alternative. This option is often technically difficult, making the proce- dure significantly more complex; furthermore, the re- sults are also unpredictable. 9 Consequently, glenoid retroversion is sometimes not, or only partially, cor- rected during total shoulder replacement. Malposition of the prosthetic components has been reported as a cause of unsatisfactory results after total shoulder arthroplasty. 8 Experimental and numeric studies have shown that misalignment of the glenoid may lead to asymmetric load of the component and cement failure. 10,17,18 The specific consequences of posterior glenoid retroversion on the survival of the implant are still not well known. Particularly, the ef- fects on micromotion at the bone-cement interface and the stress transmitted to the underlying bone have not yet been studied. The objective of this work was, therefore, to analyze, by the mean of a finite element model, the biomechanics and the risks of loosening of a glenoid implanted with different angles of retrover- sion. The amount of retroversion to correct was also evaluated. MATERIALS AND METHODS Shoulder model A 3-dimensional finite element model of the shoul- der was developed 2,3 and specifically adapted for this study. Data were obtained from an intact cadaver shoulder without any macroscopic or radiologic signs of pathology. Computed tomography sequences al- lowed reconstruction of bone shapes and density, From the a Orthopaedic Hospital, University of Lausanne and b the Orthopaedic Research Laboratory, Swiss Federal Institute of Technology Reprint requests: Alain Farron, Hôpital Orthopédique, Av. Pierre Decker 4, 1005 Lausanne, Switzerland (E-mail: alain.farron@ chuv.ch). Copyright © 2006 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2006/$32.00 doi:10.1016/j.jse.2005.10.003 521