Endocrine Abstracts

Introduction: The effects of long-term replacement therapy of primary adrenal insufficiency (PAI) are still a matter of debate. Glucocorticoid (GC) replacement regimens do not completely mimic the endogenous hormonal production and their monitoring is sometimes difficult. Therefore, some patients are exposed to mild GC excess with potential complications, such as hypertension, diabetes, and skeletal fragility. Data on bone mineral density (BMD) in PAI is still scarce and controversial.

Objective: To evaluate the impact of long-term GC treatment on BMD, in patients with PAI.

Methods: We conducted a retrospective cohort study of patients with a diagnosis of PAI, followed-up between 2011 and 2021. Patients with other causes of osteoporosis were excluded. BMD was evaluated by dual-energy X-ray absorptiometry (DXA). Doses of the various glucocorticoids were converted to hydrocortisone (HC) equivalents. We calculated cumulative doses of HC (annual and lifetime), divided by corresponding body area (mg/m2), until the date of DXA.

Results: 47 patients with PAI were included, 29 with autoimmune origin (63%), 11 with congenital adrenal hyperplasia (CAH) (23.9%), 4 from tuberculosis (8.7%), 1 from x-ALD and 1 from spontaneous bilateral bleeding (2.2%). Mean age 51.3±14.3 years, 57.5% females, disease duration 24.6±15.5 years. Mean daily HC dose was 16.6±4.3 mg/m² and mean cumulative lifetime dose was 154±139 g/m². 43 patients were under mineralocorticoid (91.5%). Osteoporosis was present in 35.7% of patients. There was an inverse correlation between cumulative lifetime GC dose and lumbar (r=−0,435, P=0,030) or femoral T-scores (r=−0.437, P=0.030); and between cumulative annual dose and lumbar T-score (r=−0.458, P=0.025). Patients with osteoporosis (any T-score ≤−2.5) had a higher cumulative lifetime HC dose (P=0.027), and a logistic regression model revealed that this association was independent of sex, PAI etiology, and treatment with mineralocorticoids (P=0.048). There was no association between T-score and type of GC replacement. There was a correlation between lumbar T-score and age (r=−0.460, P=0.016), but there was no relationship between BMD and disease duration.

Conclusion: In this study, there was an inverse linear relationship between glucocorticoid cumulative dose and bone mineral density at lumbar spine, which is in line with the known preferential action of GC on trabecular bone. Mild GC excess for some periods of life may have a greater impact in BMD reduction than disease duration. Our results reinforce the need for close monitoring of GC replacement therapy in patients with PAI.