Evaluation by digital subtraction radiography of induced changes in the bone density of the female rat mandible.

Objectives: To verify whether changes in bone mass induced by glucocorticoid and bisphosphonate can be detected by digital subtraction radiography of lateral X-rays in female rat mandibles.

Methods: 36 Rattus norvegicus (Wistar) females were randomly allocated to three groups: Group 1, control (n = 12); Group 2, induced osteopaenia (n = 12); and Group 3, induced osteopaenia plus risedronate (n = 12). Group 1 received subcutaneous saline solution (2 ml kg(-1)) injections for 12 weeks. Group 2 received 1 mg kg(-1) methylprednisolone acetate for 12 weeks. Group 3 received the same treatment as Group 2, plus 3 mg kg(-1) risedronate sodium for a further 12 weeks. A radiograph of the left mandible was taken on day 0 and after 12 weeks. A third radiograph was taken in Group 3 after risedronate treatment. Digital subtraction of the radiographs was used to compare the intensity of pixels in the control area and the test area. The animals were killed and the same area assessed by subtraction was histologically evaluated.

Results: ANOVA revealed statistically significant histological differences between the three groups. Tukey's multiple comparisons test showed that Group 2 had the smallest mean proportion of bone trabeculae per field and Group 3 the largest (F = 37.56; P < 0.01). Radiographic subtraction revealed a higher mean pixel intensity in Groups 1 and 3 vs Group 2. In Group 3, glucocorticoid caused a significant loss in radiographic density, and risedronate restored that loss (Friedman's non-parametric test).

Conclusions: Subtraction radiography was able to detect changes in bone mass induced by glucocorticoid and bisphosphonate.