Abstract
In the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, 7705 postmenopausal women with osteoporosis, defined by low bone mineral density and/or prevalent vertebral fractures (VF), were randomized to placebo or raloxifene (60 or 120 mg/day). All women received daily calcium (500 mg) and vitamin D (400–600 IU) supplements. Our previous analyses found that changes in BMD and biochemical markers of bone turnover are poorly predictive of the reduction in VF risk observed with raloxifene. This present study evaluated the effects of raloxifene on type I procollagen N-terminal propeptide (PINP), a new marker of bone turnover. Logistic regression analysis models evaluated the relationships between the changes at 1 year in PINP, serum osteocalcin (OC), bone-specific alkaline phosphatase (BSAP), and urinary excretion of type I collagen C-telopeptide fragments normalized to creatinine (CTx/Cr), and the risk of new VF at 3 years for placebo and pooled raloxifene. A subset of 967 women (mean age = 68 years) from the MORE cohort had PINP, OC, BSAP, and CTx evaluated at baseline. Both doses of raloxifene significantly decreased (
P < 0.001) all biochemical markers of bone turnover from baseline. Compared to baseline, PINP levels were decreased by medians of 11.0% and 40.8% in the placebo and pooled raloxifene groups, respectively. In addition, the placebo and pooled raloxifene groups decreased serum OC by 8.5% and 31.8%, BSAP by 15.8% and 34.6%, and urinary CTx/Cr excretion by 5.6% and 46.5%, respectively, from baseline. In the pooled raloxifene group, the logistic regression relationship between 3-year VF risk and 1-year percentage change for each biochemical marker was statistically significant with PINP (slope estimate = 0.0085,
P = 0.009), OC (slope estimate = 0.0068,
P = 0.035), and BSAP (slope estimate = 0.0056,
P = 0.039), but not with CTx/Cr (slope estimate = 0.0027,
P = 0.192). Furthermore, the percent decrease in PINP at 1 year could account for 28% of the total reduction in vertebral fracture risk. In conclusion, a 1-year decrease in PINP, BSAP, or OC, but not CTx/Cr, may be predictive of the 3-year VF risk reduction with raloxifene therapy in this subset of postmenopausal women with osteoporosis.