Influence of scapular tilt on radiographic assessment of the glenoid component after total shoulder arthroplasty: which radiographic landmarks are reliable?

Abstract

Total shoulder arthroplasty has been shown to improve function and to reduce pain in cases of osteoarthritis. To assess loosening of the glenoid component, serial evaluation of frontal plane radiographs of the scapula has been established as the "gold standard." The aim of this study was to evaluate the reliability of different bone landmarks when the scapula is tilted compared with the ideal view. Glenoid components were implanted into 6 human cadaveric scapulae. Radiographs were taken exactly anterior-posterior in the frontal plane as well as craniocaudal tilted (±15° and ±30°) and mediolateral tilted (±10° and ±20°). The following landmarks were evaluated: lateral margin of the scapula, medial margin of the scapula, floor of the fossa supraspinatus line, spine of the scapula line, glenoid fossa line, and coracoid base line. In evaluating the inclination of the glenoid component, the medial margin of the scapula had the best intraobserver and interobserver reliability with a variance for each of 2° ± 1° (P < .0001), whereas the lateral margin of the scapula had an acceptable intraobserver and interobserver reliability with a variance of 4° ± 1° and 3° ± 1°. In measuring medial migration, the glenoid fossa line had a significantly lower intraobserver and interobserver reliability than the coracoid base line (each 1 ± 0 mm vs. 3 ± 1 mm and 3 ± 2 mm; P < .0001). To assess the inclination of the glenoid component, the medial margin of the scapula has proven best, and the lateral margin of the scapula has acceptable reliability. For measuring medial migration, the coracoid base line has proven acceptable reliability, whereas the glenoid fossa line would be subject to change when osteolysis occurs at the glenoid.