Bone safety of aromatase inhibitors versus tamoxifen P.E. LØNNING Department of Medicine, Section of Oncology, Haukeland University Hospital, Bergen, Norway Abstract. Lønning PE. Bone safety of aromatase inhibitors versus tamoxifen. Int J Gynecol Cancer 2006;16(Suppl. 2):518–520. Bone loss may be a potential side effect of implementing aromatase inhibitors in the adjuvant setting. Cur- rent evidence suggests a minor bone loss during treatment with the steroidal aromatase inhibitor exemestane compared to placebo and a nonsignificant increase in fracture rate during treatment with exemestane when compared to tamoxifen. Such a difference could be due to the bone-sparing effects of tamoxifen. For the non- steroidal inhibitors letrozole and anastrozole, the MA17 study revealed a nonsignificant increase in fracture rate for letrozole compared to placebo. In contrast, both anastrozole and letrozole were found to significantly increase fracture rate compared to tamoxifen when administered as monotherapy or given sequentially. While an increased fracture rate may have detrimental effects, evidence suggests that enhanced bone loss may be preventable through careful bone mineral density (BMD) assessment and treatment with bisphosphonates. KEYWORDS: aromatase, bone metabolism, breast cancer, inhibitor. For the past two decades, tamoxifen has been the stan- dard adjuvant endocrine therapy in pre- as well as post- menopausal breast cancer patients. While the novel third-generation aromatase inhibitors given sequentially to tamoxifen (1–3) or as monotherapy (4) have improved relapse-free survival compared to tamoxifen mono- therapy, to fully assess the benefits as well as cost utility of implementing these compounds, we need to assess potential side effects as well as potential benefits on other organ systems. Apart from potential subjective side effects, the organ systems of major concern with respect to estro- gen deprivation relate to potential detrimental effects on cardiovascular risk factors (blood lipids and co- agulation factors) and bone metabolism. While large studies, like the Women Health Initiative Study (5) , could not confirm previous reports suggesting a beneficial effect of hormone replacement therapy on cardiovascular morbidity, estrogens are known to prevent bone loss in postmenopausal women (6) , and tamoxifen has been shown to have beneficial effects on bone loss as well as plasma lipids in post- menopausal women (7,8) . Effects of aromatase inhibitors on bone metabolism While bone loss is a continuous process in postmen- opausal women, it may vary substantially between different populations (9) . To assess the effect of an aromatase inhibitor on BMD, bone biomarkers, plasma lipids, and coagulation factors, we conducted a randomized placebo-controlled study (10) comparing the effect of exemestane to placebo for 2 years in a group (n ¼ 147) of patients treated for low-risk early breast cancer that would otherwise not have been sub- ject to adjuvant therapy according to the Norwegian national guidelines at that time. Here, we found exemestane to cause a slight increase in bone loss compared to placebo in the hip area but not in the lumbar spine. Interestingly, we observed an increase not only in biomarkers of bone degradation but in markers of bone synthesis as well. Most importantly, this effect on BMD as well as bone markers was partial reversible upon termination of treatment (11) . Our findings are consistent with pre- liminary data from a subprotocol of the IES study (2) , comparing BMD during treatment with exemestane to tamoxifen (12) . Following an initial drop in BMD after terminating tamoxifen, bone loss in the exemestane arm was in the same range as expected in post- menopausal women in general. Address correspondence and reprint requests to: Professor Lønning, Department of Medicine, Section of Oncology, Haukeland Hospital, Bergen, Norway. Email: per.lonning@helse-bergen.no doi:10.1111/j.1525-1438.2006.00685.x # 2006, Copyright the Authors Journal compilation # 2006, IGCS Int J Gynecol Cancer 2006, 16 (Suppl. 2), 518–520