Metabolic bone disease care practice patterns and the correlation between computed tomography and fracture risk...

Periprosthetic fracture following total shoulder arthroplasty (TSA) is a devastating complication. Metabolic bone disease has been shown to increase the incidence of fracture. However, the state of metabolic bone care in this population is poorly studied. The purpose of this study is to elucidate the state of metabolic bone care in patients undergoing TSA. We also aim to study the association between computed tomography (CT)‑derived Hounsfield unit (HU) measurements and Fracture Risk Assessment Tool (FRAX) scores in this population.

Consecutive adult patients who underwent primary TSA between January 1, 2019, and December 31, 2023, at a single institution were included in the study. Demographic variables, medical and surgical history, medications, pre-operative and post-operative dual-energy X-ray absorptiometry scans, and post-operative follow-up data were collected retrospectively. FRAX scores were calculated for each patient. Patients with FRAX major score >20 or FRAX hip score >3 were classified as "high-risk" for metabolic bone disease‑related complications. HU measurements for the humeral head and glenoid from pre-operative CT were collected for patients with available CT scans. T-tests were used to compare high-risk vs. low-risk patients. Logistic and linear regression analyses were used to examine factors associated with higher metabolic bone disease‑associated risk.

Overall, 831 patients were included in the study, with 577 of them having a pre-operative shoulder CT scan. A total of 632 patients were classified as low risk, whereas 199 were classified as high risk. For high-risk patients, 50% and 16% received pre-operative and 1-year post-operative dual-energy X-ray absorptiometry, respectively. Of these, 60% were on vitamin D and calcium, and only 13% were taking osteoporosis medications. High-risk patients were significantly more likely to be older, female, obese, and to have a higher comorbidity index. The average humeral head HU measurement was significantly lower in high-risk patients (28.5 HU vs. 50.14 HU, P < .001) and in patients with diagnosed osteoporosis (23.6 HU vs. 44.4 HU, P = .005). Univariate regression demonstrated a strong correlation between humeral HU measurement and metabolic bone disease risk.

Metabolic bone disease is under-screened and under-treated in patients undergoing shoulder arthroplasty, with a large proportion of high-risk patients not receiving indicated care. There are significant missed opportunities to improve metabolic bone care for these patients during the peri-operative period, indicating barriers to care. High-risk patients and patients with diagnosed osteoporosis have significantly lower humeral HU on pre-operative CT.