Correlation between radiographic risk for glenoid component loosening and clinical scores in shoulder arthroplasty

Glenoid component loosening is the weak point in the failure of total shoulder arthroplasty (TSA). In this study we analyse the radiographic risk factors observed on 86 cemented polyethylene glenoid components and their relationship with clinical signs at a mean follow-up of 5.8 years. Clinical assessment included Simple Shoulder Test (SST) and Constant-Murley score. Radiograms were taken to detect periprosthetic radiolucency, tilt, medial displacement and polyethylene thinning. Pearson's correlation coefficient and Spearman's rank correlation coefficient were calculated for statistical analysis. In 61 patients (71%) lucent lines were less than 2 mm wide (grade 2) and in 6 cases (7%) they were >or=2 mm wide (grade 3 and 4). Thinning of the polyethylene was found in 11 cases (13%), glenoid tilt in 6 cases (7%) and medial migration of the component in 5 cases (6%). Complete glenoid prosthetic loosening was found in 3 cases (3.5%) associated with polyethylene wear and glenoid bone loss. The Constant-Murley score associated with radiolucency grade 3 and 4 was less than 45% (38.39 +/- 8.9) (p < 0.05), while a score less than 56% (30.72 +/- 8.7) was found in patients with glenoid tilt and medial migration (p < 0.01). The mean SST score was 4.8 +/- 2.8 in case of glenoid tilt and migration of the component (p < 0.01). Removal of the glenoid component and conversion to hemiarthroplasty or reverse prostheses is suggested in painful glenoid loosening. An exhaustive analysis of radiograms is essential to detect early and late complications or risk factors of glenoid loosening.