Breast cancer, osteoporosis, and coronary heart disease (CHD) are three major threats to women’s health. Postmenopausal women with breast cancer are also at high risk for osteoporosis and CHD. Adjuvant tamoxifen therapy is not only an effective treatment for breast cancer, but has been shown to have a beneficial effect on bone and the cardiovascular system. Although tamoxifen has been convincingly demonstrated to be able to preserve bone mineral density (BMD), an unexpected increase of risk of fractures in patients treated with tamoxifen has been reported. The findings of the association of tamoxifen and CHD from previous studies were either borderline or inconsistent. To clarify the discrepancy between BMD and fractures and test the potential beneficial effect of tamoxifen on CHD, I conducted a series of retrospective studies in postmenopausal women with breast cancer who participated in the Cancer Surveillance in HMO Administrative Data (IMPACT study) or the Study of Osteoporotic Fractures (SOF).
In patients who participated in the IMPACT study, I demonstrated that the association of tamoxifen and fracture incidence varied at different skeletal sites. Although the association of tamoxifen and fractures in the spine (HR=0.40, 95% CI: 0.09-1.85), wrist (HR=2.49, 95% CI: 0.88-7.06), and total body (HR=0.87, 95% CI: 0.49-1.55) was inconclusive, tamoxifen was associated with an apparent reduction of the risk of hip fracture (HR=0.41, 95% CI: 0.17-1.03, p=0.0565). Importantly, the pattern of observed association of tamoxifen with the risks of fractures among postmenopausal women with breast cancer is consistent with its widely reported preserving effect on bone mineral density.
Using SOF data, I found that the association between BMD and fractures in women with breast cancer varied at different skeletal sites, and type of BMD measured. Non-specific BMD was not associated with hip fracture (HR=1.12; 95% CI: 0.78, 1.59). Site-specific BMD was more likely linked with hip fracture (HR=1.43, 95% CI: 0.99, 2.08) while change in BMD did not predict hip fracture (HR=1.05; 95% CI: 0.63, 1.72). The association of spine morphometric fracture with either non-specific or spine-specific BMD was similar (OR=1.40; 95% CI: 1.04, 1.90; OR=1.35, 95% CI: 0.99, 1.85, respectively). Overall, the association of BMD and fracture in elderly women with breast cancer is weak. Only site-specific BMD appears to have a consistently modest association with fractures in the corresponding skeletal sites.
In the IMPACT study population, compared to patients without tamoxifen, the overall incidence of CHD in tamoxifen-treated patients was lower (adjusted HR=0.60, 95% CI: 0.40-0.88). For each year of tamoxifen use, there was a statistically significant decrease in the risk of CHD (HR=0.90, 95% CI: 0.82-0.98). Further analyses categorized by length of tamoxifen use showed that an apparent association with a decreased CHD risk was found in patients who received tamoxifen for two to five years (HR=0.54, 95% CI: 0.33-0.86). No association was detected after the discontinuation of tamoxifen therapy.
In summary, I detected a possible benefit associated with tamoxifen on fractures in the hip, the most common fracture site. I also found that BMD did not predict osteoporotic fractures well in postmenopausal women with breast cancer. In addition, I demonstrated that tamoxifen was associated with a reduced risk of CHD in postmenopausal women with breast cancer in a dose-dependent manner. An apparent benefit was found in those patients who received tamoxifen therapy for at least two years.
| Identifer | oai:union.ndltd.org:umassmed.edu/oai:escholarship.umassmed.edu:gsbs_diss-1405 |
| Date | 28 December 2008 |
| Creators | Ding, Hongliu |
| Publisher | eScholarship@UMassChan |
| Source Sets | University of Massachusetts Medical School |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Morningside Graduate School of Biomedical Sciences Dissertations and Theses |
| Rights | Copyright is held by the author, with all rights reserved., select |
Page generated in 0.0021 seconds