Impacts of aromatase inhibitors on site-specific bone mass and geometry in postmenopausal women with endocrine-responsive breast cancer

Abstract

The negative impact of aromatase inhibitor (AI) used as adjuvant hormonal therapy for postmenopausal women with breast cancer on bone mass determined by dual energy X-ray absorptiometry and increased risk of fragility fractures are obvious. As fracture risk highly depends on bone structures as well as bone mass, we have adapted quantitative computed tomography (QCT) as an effective way of evaluating both volumetric bone mineral density (vBMD) and geometric parameters in bones in breast cancer patients treated with AI.To analyze the change in bone geometry caused by AI, 375 patients in their fifties diagnosed as breast cancer were chosen. The subjects who received AI (n=196, mean age 54.3 ± 0.2 y.o.) and the ones who did not get any adjuvant therapy (n=179, mean age 53.8 ± 0.2 y.o.) had been examined by QCT at the time point of average 23.6 ± 1.1 months after the mastectomy. Volumetric (3-dimensional) and areal (2-dimensional) bone mineral density (BMD) and hip geometry were analyzed by QCT PRO (Mindways Software, Austin TX, USA). There were no statistical differences in volumetric BMD of lumbar spines. Areal BMD in total hip (TH) was significantly lower in patients who had been treated with AI compared to non-AI treated group (0.511 ± 0.008 vs. 0.488 ± 0.007 g/cm2, p <0.01). Cross-sectional area of femur neck, trochanter, intertrochanter were significantly lower in the AI group (1.92 ± 0.04 vs. 1.81 ± 0.02 cm2, 1.7 ± 0.03 vs. 1.61 ± 0.25 cm2, 2.79 ± 0.04 vs. 2.15 ± 0.03 cm2, 2.27 ± 0.04 vs. 2.15 ± 0.03 cm2; all p <0.05, respectively) Cortical thickness was decreased in every part of the femur, femoral neck, trochanter, intertrochanter in the AI group (2.27 ± 0.05 vs. 2.10 ± 0.04 mm, 1.62 ± 0.05 vs. 1.48 ± 0.04 mm, 3.62 ± 0.05 vs. 3.49 ± 0.04 mm, 2.7 ± 0.04 vs. 2.55 ± 0.04 ; all p <0.05, respectively). Cortical buckling ratio tended to increase in

the AI group, only significantly higher in femur neck part (7.12 ± 0.19 vs. 7.90 ± 0.35, p=0.05). There were no differences regarding angle, width, axis length between two groups. These findings indicate that AI leads to changes in the increased endocortical bone resorption with accordingly reduced cortical thickness in the femur earlier than expected before the occurence of the well-known bone loss in the spines.