Article Text

SAT0472 The Effect of Aromatase Inhibitors on Different Sites of the Human Skeleton – an Observational Case-Control Study
  1. J. Fowler1,2,
  2. K. Mitchell1,2,
  3. M. Bukhari1,3
  1. 1Faculty of Health and Medicine, Lancaster University
  2. 2University Hospitals of Morecambe Bay, Lancaster, United Kingdom
  3. 3Rheumatology department, University Hospitals of Morecambe Bay, Lancaster, United Kingdom


Background Evidence suggests that patients on aromatase inhibitors (AIs) for breast cancer treatment are at an increased risk of bone loss therefore during AI treatment they continuously have their bone mineral density (BMD) monitored (1). Hip BMD is used to measure the future fracture risk as recommended by FRAX. The MAP.3 randomized controlled study found that patients on exemestane, which is an AI, lost cortical BMD at a quicker rate than trabecular BMD (2). This is the only study to find this pattern of bone loss in the human skeleton.

Objectives The aim of this study was to establish the pattern of BMD loss in patients on aromatase inhibitors using vertebral bone as a surrogate for trabecular bone and the hip as a surrogate for cortical bone in patients undergoing treatment and having multiple scans.

Methods Using data from a district dual energy x-ray absorptiometry (DEXA) from the North West of England, a nested case-control study was designed. A cohort of patients referred between 2004 and 2011 who had received AI treatment as their main risk factor and who had been scanned more than once were identified. These were then age and gender matched with controls referred in the same time period with no indication for scanning and who were scanned more than once with the same age of initial scanning as the cases. Sites of bone loss were ascertained by comparing the BMD in two areas (L1-L4 vertebrae and the left femoral neck) between cases and controls. A logistic model was then fitted to determine the odds of having a significantly reduced BMD with AI treatment which was then adjusted to body mass index (BMI) of the cohort of cases and controls. This was done for both sites. The fit of the model was tested using receiver operating characteristic (ROC) curves.

Results 96 female patients on AI treatment were identified, with a median age of 63.9 years (IQR 59.2, 72.4) at the first scan. 96 controls were age and sex matched for the first scan. Median time between scans was 3.3 years (IQR 3.0, 4.1). The unadjusted odds of a lower BMD were 22.8 (CI 95% 2.1, 238.2) for the femoral neck and 5.8 (CI 95% 1.4, 24.3) for the lumbar spine, repectively. The odds of a lower BMD when adjusted for BMI were 27.7 (95% CI 2.6 – 301.7) for the femoral neck and was a lot higher than the odds ratio of 6.5 (95% CI 1.5 - 28.1) found in the lumbar spine. The adjusted area under the curve was 0.628 in the femoral neck and 0.6404 in the lumbar spine.

Conclusions Both femoral neck BMD and lumbar spine BMD are significantly reduced over time in patients on aromatase inhibitors. The chance of having a significant amount of bone loss is a lot higher in the cortical bone of the femoral neck than the trabecular bone of the lumbar spine. AIs have more of an effect on cortical bone than trabecular bone by making the cortical bone more porous. This increases the risk of fractures in cortical bone (2). Since hip BMD is a predictor of future fracture risk, patients on AIs need to be monitored intensely.


  1. Official journal of the European Society for Medical Oncology/ESMO. 2011;22(12):2546-55. Epub 2011/03/19.

  2. The New England journal of medicine. 2011;364(25):2381-91. Epub 2011/06/07.

Acknowledgements None.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.2225

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.