Objectives:​ To compare the efficacy of denosumab and alendronate on raising spine bone mineral density (BMD) in long-term glucocorticoid (GC) users.

Methods:​ Adult patients receiving long-term prednisolone (≥2.5 mg/day for ≥1 year) were recruited and randomized to either subcutaneous denosumab (60 mg/6 months) or oral alendronate (70 mg/week). BMD (lumbar spine, femoral neck, hip) and bone markers (serum P1NP and CTX) were measured at month 0, 6 and 12. The difference in spine BMD (primary outcome) at month 12 was compared between the two groups.

Results:​ 139 subjects were recruited (age 50.0 ± 12.7 years;​ 96% women):​ 69 assigned denosumab and 70 assigned alendronate. At entry, 73(53%) patients were osteoporotic and 82(59%) patients were naive to the bisphosphonates. Baseline clinical characteristics and BMD values were similar in the two groups. At month 12, a significant gain in mean BMD at the lumbar spine (+3.5 ± 2.5%;​ p<0.001), hip (+0.9 ± 2.8%;​ p=0.01) and femoral neck (+1.04 ± 4.1%;​ p=0.047);​ was observed in denosumab-treated patients, whereas the corresponding change was +2.5 ± 2.9% (p<0.001), +1.6 ± 2.7% (p<0.001) and + 1.5 ± 3.9% (p=0.002) in the alendronate group. The spine, but not the hip or femoral neck, BMD at month 12 was significantly higher in the denosumab than alendronate group after adjustment for baseline BMD values, age, sex, osteoporosis risk factors and the cumulative prednisolone doses received in one year. The drop in P1NP and CTX was significantly higher in the denosumab than alendronate group. Frequency of adverse events (AEs), including infections, was similar in the two treatment arms. Seven patients withdrew from the study but not related to AEs.

Conclusions:​ In patients receiving long-term GCs, denosumab is superior to alendronate in raising the spine BMD after 12 months. Both drugs are well-tolerated.