MINERAL DENSITY GAIN IN VERTEBRAE OF OSTEOPOROTIC WOMEN ON ORAL PAMIDRONATE REVERTS A YEAR AFTER TREATMENT DISCONTINUANCE

Long-term treatment with 200 mg dry weight/day of pamidronate (APD) by oral route has been proposed to increase bone mineral density (BMD) i n postmenopausal osteoporotic women. However, there is widespread conc ern over the possibility of bone metabolism ''freezing'' by protracted use of medication liable to inhibit bone resorption. Accordingly, an open population of 39 osteoporotic women was studied in order to deter mine BMD variations in lumbar vertebrae and femoral neck and to confir m whether given APD doses increase bone mineralization. A parallel gro up of osteoporotic women was likewise evaluated to determine the poten tial reversibility of such effect on discontinuing treatment. Results were beneficial at both skeletal sites in subjects on APD therapy, dis closing at 18 months treatment 9.0% mean peak mineral gain versus basa l status at lumbar spine. In the few responders completing 4 years tre atment, mean differences versus basal status were +4.9% in vertebrae a nd +6.2% at femoral neck. In lumbar vertebrae there was a rapid trend in mineral gain up to 18 months on treatment, which declined thereafte r to a quarter of its early rate. Concurrently, in the group of 21 bas eline-matched women, effects were evaluated after discontinuing daily APD administration one year (n = 11) or two years (n = 7) later. In bo th subgroups, there was a loss in lumbar (-7.1% and -9.8% respectively ; p < 0.01) and femoral neck BMD (-2.2% and -4.2% respectively; p: n.s .) on performing measurement one year after APD withdrawal. Furthermor e, after three years treatment and subsequent discontinuance, three pa tients presented bone loss in lumbar vertebrae and minimal changes at femoral neck one year after APD withdrawal. Therefore, APD induces mod erate BMD gain which proves reversible on discontinuing therapy, so th at it seems unlikely that this compound should ''freeze'' bone metabol ism to a significant extent. However, the precise degree of such rever sibility requires evaluation in larger series.