The effect of raloxifene on bone marrow adipose tissue and bone turnover in postmenopausal women with osteoporosis.

In patients with postmenopausal osteoporosis low bone volume is associated with high bone marrow adipose tissue (MAT). Moreover, high MAT is associated with increased fracture risk. This suggests an interaction between MAT and bone turnover, however literature remains equivocal. Estrogen treatment decreases MAT, but the effect of raloxifene, a selective estrogen receptor modulator (SERM) registered for treatment of postmenopausal osteoporosis, on MAT is not known. The aim of this study is 1] to determine the effect of raloxifene on MAT and 2] to determine the relationship between MAT and bone turnover in patients with osteoporosis. Bone biopsies from the MORE trial were analyzed. The MORE trial investigated the effects of raloxifene 60 or 120mg per day versus placebo on bone metabolism and fracture incidence in patients with postmenopausal osteoporosis. We quantified MAT in iliac crest biopsies obtained at baseline and after 2years of treatment (n=53; age 68.2±6.2years). Raloxifene did not affect the change in MAT volume after 2years compared to baseline (placebo: 1.89±10.84%, raloxifene 60mg: 6.31±7.22%, raloxifene 120mg: -0.77±10.72%), nor affected change in mean adipocyte size (placebo: 1.45 (4.45) μm, raloxifene 60mg: 1.45 (4.35) μm, raloxifene 120mg: 0.81 (5.21) μm). Adipocyte number tended to decrease after placebo treatment (-9.92 (42.88) cells/mm) and tended to increase during raloxifene 60mg treatment (13.27 (66.14) cells/mm) while adipocyte number remained unchanged in the raloxifene 120mg group, compared to placebo (3.06 (39.80) cells/mm, Kruskal-Wallis p=0.055, post hoc: placebo vs raloxifene 60mg p=0.017). MAT volume and adipocyte size were negatively associated with osteoclast number at baseline (R=0.123, p=0.006 and R=0.098, p=0.016 respectively). Furthermore adipocyte size was negatively associated with osteoid surface (R=0.067, p=0.049). Finally, patients with vertebral fractures had higher MAT volume (50.82 (8.80)%) and larger adipocytes (55.75 (3.14) μm) compared to patients without fractures (45.58 (12.72)% p=0.032, 52.77 (3.73) μm p=0.004 respectively). In conclusion, raloxifene did not affect marrow adipose tissue, but tended to increase adipocyte number compared to placebo. At baseline MAT volume and adipocyte size were associated with bone resorption, and adipocyte size was associated with osteoid surface, suggesting an interaction between bone marrow adipocytes and bone turnover. In addition, we found that high MAT volume and larger adipocyte size are associated with prevalent vertebral fractures in postmenopausal women with osteoporosis, indicating that adipocyte size affects bone quality independent of bone volume.