Radiographic sclerotic contour loss in the identification of glenoid bone loss.

Abstract

Quantification of glenoid bone loss guides surgical management in the setting of anterior shoulder instability. Glenoid defects resulting in ≥20% articular area loss require bony reconstruction. The objective of this study was to evaluate the utility of sclerotic glenoid contour loss on true anteroposterior radiography in the detection of varying quantities of simulated glenoid bone loss using a cadaveric model. Eight cadaveric scapulae with full radiographic sclerotic contour were osteotomized to produce glenoid surface area reductions of 10-50%. Radiography was performed initially and following each osteotomy, and assessed by an orthopedic surgeon and radiologist twice. Quantity of glenoid loss was compared using Fisher's exact test. Sensitivity, specificity, and reliability analyses were performed. On the first radiographic review, sclerotic contour loss was detected in 6 out of 8 scapulae with 50% area loss, but only 1 out of 8 scapulae with 20% area loss. There was a significantly higher proportion of radiographs containing sclerotic contour loss for defects with 50% area loss compared to those with 0-25% loss (p≤0.02). In the detection of ≥20% area loss, sclerotic contour loss had a sensitivity of 33-43% and specificity of 88-100%. Moderate inter-observer reliability (Cohen's kappa value of 0.42-0.53) and intra-observer reliability (kappa value of 0.46-0.58) were found. Radiographic sclerotic contour loss is commonly observed in radiographs of scapulae with 40-50% glenoid area loss and less often with smaller lesions. However, this finding lacks utility in discerning specific quantifications of glenoid bone loss. In a clinical setting, sclerotic contour loss suggests the presence of a large glenoid defect that may require bony reconstruction. However, an intact sclerotic contour does not rule out significant bone loss.