Predicting subsequent bone density response to intermittent cyclical therapy with etidronate from initial density response in patients with osteoporosis.

We investigated whether an increase in lumbar spine bone mineral density (LS BMD) at 6 months or at 12 months could predict the response to intermittent cyclical therapy (ICT) with etidronate, defined in one of two ways: (i) an increase in LS BMD at 24 months (improvement) or (ii) an increase in LS BMD > or = 0.028 g/cm2 (significant improvement). The latter is a precision term calculated from test-retest values for LS BMD in osteoporotic patients. Two hundred and forty-seven patients (32 men; 5 premenopausal and 210 postmenopausal women) were followed for 24 months by dual-energy X-ray absorptiometry (DXA) and were not taking estrogen, calcitonin or fluoride during treatment with ICT-etidronate. One hundred and fifty patients had a LS BMD measurement after 6 months of treatment with ICT-etidronate and 205 patients had one at 12 months. Baseline characteristics (mean;SD) were as follows: age, 66;11 years; years since menopause, 21;10; number of vertebral fractures at baseline, 0.87;1.26; LS BMD T-score, -2.8; 1.2. After 24 months of treatment with ICT-etidronate, 81% of the patients had an improvement, and 55% had a significant improvement at the LS. Only 6% significantly lost bone (loss of 0.028 g/cm2 or more). The mean percent change from baseline in LS BMD was 5.1% (95% confidence interval 4.2% to 6.0%). The results for men and postmenopausal women were similar to those for the entire group. Accuracy and sensitivity were marginally, but not significantly, higher when response was predicted using 12 month versus 6 month LS BMD measurements. The positive predictive values of improvement at 6 or 12 months were 89% and 90% respectively for improvement at 24 months, and 66% and 68% for significant improvement at 24 months. Identification of nonresponders was less successful and similar at 6 months and 12 months. Forty percent and 39% of the patients, who had no improvement at 6 or 12 months respectively, also had no improvement at 24 months, i.e., were true negatives, while 77% and 71% had no significant improvement at 24 months. The results may reflect slow response in a small subgroup of patients rather than nonresponse; however, no response at 1 year might identify patients whose rate of response is sufficiently slow that alternative therapy is justified. These data demonstrate a good response rate to ICT-etidronate and may help reduce the need for follow-up BMD measurements in those who show an early improvement.