Trabecular bone score as one of the new methods of non-invasive evaluation of bone microarchitecture in patients with Cushing’s syndrome

Objective - to evaluate the value of trabecular bone score and risk factors of fractures in patients with Cushing’s syndrome (CS). Material and methods. One hundred eighty two patients with laboratory-confirmed Cushing’s syndrome were enrolled. All patients underwent measurement of bone mineral density (BMD) at the lumbar spine (LI-LIV), femoral neck and total hip using DXA Prodigy (GEHC Lunar, Madison, WI, USA). Trabecular bone score (TBS) was assessed retrospectively on the basis of already existing DXA images using software TBS iNsight software v2.1 (Medimaps, Merignac, France). Each patient was interviewed for the presence of low-traumatic fractures during the active stage of the disease. A lateral X-ray of the thoracic and lumbar spine ThIV-LV was performed to estimate vertebral fractures. Twenty-four hours urinary free cortisol (24hUFC) was measured by imunochemiluminescence assay VITROS ECi with the preliminary extraction with diethyl ether (reference values 60-413 nmol/24 h). Results. Among 182 patients with CS (149 women, 33 men), Cushing’s disease was confirmed in 151 cases, 9 patients diagnosed with benign adrenal tumor and 22 - ACTH-ectopic syndrome. The median of age - 35 (Q25-Q75 27-49) years, body mass index - 29 (26-33) kg/m2, 24hUFC - 1760 (985-2971) nmol/24h. Fractures were confirmed in 80 (44%) cases, 70 patients suffered from vertebral fractures, which were multiple in 53 cases; 23 patients had non-vertebral fractures. Median of trabecular bone score was 1.205 (1.102-1.307), which is much lower than expected in healthy volunteers (>1.350), while the decrease in bone mineral density (BMD) did not correspond to the severity and prevalence of osteoporosis fractures: LI-LIV Z-score - 1.7 (2.5-0.73); femoral neck Z-score - 1 (-1.6- -0.4). However, when using binary logistic regression analysis (adjusted for sex, age, body mass index, bone mineral density, trabecular bone score and 24hUFC) revealed that the most significant predictor of fracture is high levels of 24hUFC (p=0.001) . The high prevalence of vertebral fractures in patients with CS most likely influenced the results of trabecular bone score and its ability to predict fractures. Conclusion. Patients with active CS have low trabecular bone score, rather than bone mineral density, which reflects deterioration in bone microarchitecture. The low-traumatic fracture occurrence depends on the severity of CS reflected in 24hUFC levels.