Hormone replacement therapy after a diagnosis of breast cancer in relationto recurrence and mortality

Background: Hormone replacement therapy (HRT) is typically avoided for wome n with a history of breast cancer because of concerns that estrogen will st imulate recurrence. in this study, we sought to evaluate the impact of HRT on recurrence and mortality after a diagnosis of breast cancer. Methods: Da ta were assembled from 2755 women aged 35-74 years who were diagnosed with incident invasive breast cancer while they were enrolled in a large health maintenance organization from 1977 through 1994, Pharmacy data identified 1 74 users of HRT after diagnosis. Each HRT user was matched to four randomly selected nonusers of HRT with similar age, disease stage, and year of diag nosis. Women in the analysis were recurrence free at MRT initiation or the equivalent time since diagnosis. Rates of recurrence and death through 1996 were calculated. Adjusted relative risks were estimated by use of the Cox regression model. All statistical tests were two-sided. Results: The rate o f breast cancer recurrence was 17 per 1000 person-years in women who used H RT after diagnosis and 30 per 1000 person-years in nonusers (adjusted relat ive risk for users compared with nonusers = 0.50; 95% confidence interval [ CT] = 0.30 to 0.85), Breast cancer mortality rates were five per 1000 perso n-years in HRT users and 15 per 1000 person-years in nonusers (adjusted rel ative risk = 0.34; 95% CI = 0.13 to 0.91). Total mortality rates were 16 pe r 1000 person-years in HRT users and 30 per 1000 person-years in nonusers ( adjusted relative risk = 0.48; 95% CI = 0.29 to 0.78). The relatively low r ates of recurrence and death were observed in women who used any type of HR T (oral only = 41% of HRT users; vaginal only = 43%; both oral and vaginal = 16%). No trend toward lower relative risks was observed with increased do se. Conclusion: We observed lower risks of recurrence and mortality in wome n who used HRT after breast cancer diagnosis than in women who did not. Alt hough residual confounding may exist, the results suggest that HRT after br east cancer has no adverse impact on recurrence and mortality.