SAT-805 Romosozumab Increases Spine and Hip Bone Mineral Density in Women With Anorexia Nervosa

Abstract Disclosure: E. Meenaghan: None. Z. Joseph: None. S. Wasserman: None. M. Slattery: None. J. Rosenblum: None. R.J. Keane: None. K.T. Eddy: None. H. Lee: None. Anorexia nervosa is complicated by low bone density (BMD) and increased fracture risk, but no pharmacologic approaches have increased BMD >6% on average nor normalized BMD in a significant number of women. The aim of this pilot study was to determine whether off-label administration of romosozumab, a humanized monoclonal antibody that targets sclerostin and addresses both the low bone formation and high bone resorption observed in anorexia nervosa, would increase BMD in women with this disorder. We conducted a 12-month, randomized, placebo-controlled study of 30 ambulatory women with low-weight or atypical (normal-weight) anorexia nervosa, ages 20-60 y, with low BMD (Z-score <-1.0) at the spine, hip, and/or radius. Participants were randomized 2:1 to 12 months of romosozumab (210 mg subcutaneously q4 weeks) (n=20) or placebo (n=10). All participants received calcium 1000 mg and vitamin D 800 IU daily. The primary outcome measure was % change in postero-anterior (PA) lumbar spine BMD by dual-energy x-ray absorptiometry (DXA). Secondary outcome measures included % change in lateral spine, total hip, femoral neck, and total radius BMD, the bone formation marker procollagen type I N-terminal propeptide (P1NP), and the bone resorption marker C-terminal telopeptide (CTX). Estrogen status was defined as “estrogen replete” if eumenorrheic or on exogenous estrogen, “estrogen non-replete” if amenorrheic not secondary to contraception, or “unknown.” Statistical analysis was performed using a linear random effects model using 0, 6 and 12-month data. At baseline, mean age [34 ± 13 [romosozumab (ROMO)] vs 27 ± 6 y [placebo (PBO)] (mean ± SD)], BMI (19.2 ± 2.2 vs 18.8 ± 1.2 kg/m2), and BMD (PA spine Z-score -1.6 ± 0.9 vs -2.0 ± 0.6) were similar between groups. Ninety percent (n=27/30) of subjects had PA lumbar spine BMD Z-score < -1. At 12 months, PA lumbar spine BMD increased in the ROMO vs PBO group (p<0.0001), with a mean increase of 12.4 ± 4.2% in the ROMO group vs a decrease of 0.3 ± 3.7% in the PBO group. In the ROMO group, mean PA lumbar spine BMD Z-score was -0.4 ± 0.9 at 12 months, and the percent of subjects with PA lumbar spine BMD Z-score ≤ -1.0 decreased to 16% (n=3/19) from 85% (n=17/20). Lateral lumbar spine (16.1 ± 6.4 vs -0.5 ± 4.6%, p<0.0001), total hip (4.1 ± 3.7 vs 1.1 ± 2.6%, p=0.005), and femoral neck (4.1 ± 3.7 vs -0.4 ± 4.5%, p=0.003), but not radius, BMD also increased in the ROMO vs PBO group. Neither baseline BMD, age, estrogen status, nor change in BMI predicted change in BMD at any site. At 3 months, serum P1NP (78 ± 135% vs -5 ± 31%, p=0.01) increased in the ROMO vs PBO group; there was no change in serum P1NP over 6 or 12 months, nor in CTX at any timepoint. Romosozumab was well tolerated. In conclusion, in this pilot, randomized, placebo-controlled study, 12 months of therapy with romosozumab markedly increased spine and hip BMD in women with anorexia nervosa, increasing BMD Z-score to >-1 in a significant number of women. Presentation: Saturday, July 12, 2025