Tamoxifen discontinuance among women aged 65 years and older with estrogen receptor positive breast cancer

Abstract

648 Background: Five years ofadjuvant tamoxifen therapy in estrogen receptor (ER)-positive breast cancer is more effective than two years of use. However information on adjuvant tamoxifen therapy discontinuance is scanty. We aimed to identify predictors of tamoxifen discontinuance in older women with early stage breast cancer. Methods: Within six healthcare delivery systems, we identified 1080 women ≥65 years with stage I-IIb breast cancer diagnosed from 1990 to 1994 with ER-positive or indeterminate tumors who had been prescribed tamoxifen. We then followed them for five years post initial tamoxifen prescription. Data sources included tumor registries, administrative data, and medical records. We used automated pharmacy records to validate tamoxifen prescription information collected from medical records. The primary endpoint was tamoxifen discontinuance, operationalized as ever discontinuing tamoxifen during follow-up. We excluded women who discontinued tamoxifen due to recurrence, death or disenrollment from this definition. Independent variables included demographic characteristics, comorbidity, tumor characteristics, and primary therapy. We used Cox proportional hazards to determine predictors of tamoxifen discontinuance. Results: Of the 1080 women who were prescribed tamoxifen, 974 women did not have a recurrence, die or disenroll during follow-up. Of these 974 women, 489 (50%) stopped taking tamoxifen during the five years of follow-up. Discontinuers were more likely to be women aged 80 years and older (versus aged 65–69, hazard ratio (HR)= 1.96, 95% confidence interval (CI)=1.51–2.56); have a Charlson Comorbidity Index of 1 (versus a Charlson Comorbidity Index of 0, HR=1.31, CI=1.07–1.61); have indeterminate receptor status (versus ER-positive, HR=1.49, CI=1.12–2.00); and to have not received breast-conserving surgery with radiation therapy or mastectomy (HR=1.37, CI=1.10–1.71). Conclusion: Attention to non-adherence among women at risk of discontinuance, particularly not receiving guideline care, might improve breast cancer outcomes for these women. No significant financial relationships to disclose.