Introduction: Since the soluble receptor activator of the NF-kappaB ligand (sRANKL) as well as the endogenous anti-resorptive cytokine osteoprotegerin (OPG) are produced by osteoblasts and given that these cells undergo significant changes during antiresorptive treatment, we hypothesized that treatment with bisphosphonates (BP) would be accompanied by changes in serum OPG and sRANKL levels.
Methods: In a prospective, randomized controlled trial of previously untreated postmenopausal women with osteoporosis, oral BP therapy (daily doses of either 10 mg alendronate or 5 mg risedronate) in combination with calcium/vitamin D was compared to calcium/vitamin D treatment alone (control group). Follow-up at 2, 6 and 12 months was completed for 56 patients. Standardized spinal X-rays were performed at baseline, and DEXA measurements at the femoral neck and trochanter were made at baseline and after 1 year. Serum OPG and sRANKL levels were measured with a polyclonal antibody-based ELISA system.
Results: After 1 year, there was a non-significant loss in neck and trochanteric bone mineral density (BMD) in the CTR group and a mean increase of 3.3% and 4.6% in the combined BP group (both p<0.0001), respectively. Serum levels of C-terminal telopeptides of type I collagen (sCTX) and osteocalcin decreased by 12% and 10% at 12 months in the CTR group and by 43% and 23% in the combined BP group, respectively (all significant). OPG serum levels in the CTR group decreased significantly by 9% at 2 months (p<0.005) and remained below pre-treatment levels at later time points. Both the alendronate- and risedronate-treated patient groups showed unaltered OPG levels after 2 months, but they had significantly increased serum levels at 6 and 12 months. Levels of sRANKL were unchanged throughout the treatment period. Univariate regression analysis demonstrated that changes in serum OPG levels after 12 months of BP treatment were positively and better correlated to BMD changes (trochanter: r=0.59, p<0.0001; neck: r=0.50, p<0.001) than those of sCTX, which showed the expected negative correlation to BMD change (trochanter: r=-0.35, p=0.03; neck: r=-0.23, p=0.16). With multiple regression analyses at 12 months, R2 values for 1-year changes in trochanteric BMD of 0.33 (OPG alone) and 0.23 (sCTX alone) were significantly improved to the 0.57 when OPG and sCTX changes were combined (p<0.001). Results for the femoral neck were also statistically significant R2=0.35, p<0.001). BMD and OPG changes in the CTR group were not correlated with each other.
Conclusions: We conclude that with BP treatment, changes in serum OPG levels, unlike changes in sCTX levels, are positively correlated to changes in BMD response. The BP-related changes in serum OPG levels during treatment could result from effects on osteoclastogenesis and osteoclast apoptosis as well as from a direct stimulatory effect on osteoblastic OPG production. These changes in OPG levels may be used to predict the individual response of patients to BP treatment.