Preoperative glenoid bone density is associated with systemic osteoporosis in primary shoulder arthroplasty

Abstract

BackgroundIn shoulder arthroplasty, poor bone mineral density can impact humeral and glenoid fixation, as well as increase the risk of periprosthetic fracture. The purposes of this study were to (1) determine the association between preoperative glenoid and humeral Hounsfield units (HUs) with systemic bone density in primary shoulder arthroplasty, (2) determine a threshold for HU on preoperative shoulder computed tomography (CT) to classify patients as osteoporotic, and (3) identify which anatomic T-score locations were significantly associated with glenoid and/or humeral HU measurements. MethodsConsecutive patients who had both a preoperative shoulder CT and a dual-energy x-ray absorptiometry (DEXA) scan were included. HU measurements of the glenoid vault and proximal humerus were recorded. DEXA T-scores from the spine, hip, and femoral neck were recorded, and osteoporosis was defined as a T-score of −2.5 or lower in any of the anatomic sites. Patients were grouped into osteoporosis and nonosteoporosis cohorts and these cohorts were compared using standard statistical tests. To assess the association between HUs and DEXA T-score, linear regression analysis was performed. Additional receiver operating characteristic analysis was performed to evaluate the ability of HUs to classify patients as osteoporotic. ResultsA total of 162 patients undergoing primary shoulder arthroplasty with both preoperative CT and DEXA scores were included. The osteoporosis cohort had significantly lower body mass index (P = .001), although no other significant differences in demographics were found between the 2 cohorts. The osteoporosis cohort was associated with significantly lower glenoid vault bone density in HUs (183 ± 104 vs. 303 ± 157, P = .004) and trended toward significantly lower proximal humeral bone density in HUs (34.3 ± 54.8 vs. 64.5 ± 58.2, P = .058). Glenoid HU threshold of 175 had an area under the curve of 0.75, with sensitivity of 80.8% and specificity of 62.5% for classifying systemic osteoporosis. HUs of both the glenoid (P = .031) and proximal humerus (P = .003) were significantly associated with femoral neck T-scores on linear regression analysis. Similarly, HUs of both the glenoid (P < .001) and proximal humerus (P = .009) were significantly associated with spine T-scores on linear regression analysis. DiscussionThis study found that glenoid vault HU is significantly associated with systemic osteoporosis and can be used as a screening tool for detecting osteoporosis preoperatively. In addition to implications for component fixation and periprosthetic fracture, local bone density evaluation on shoulder CT prior to shoulder arthroplasty represents an opportunity for osteoporosis screening.