Increased incidence of small and well-differentiated breast tumours in post-menopausal women following hormone-replacement therapy

Exposure to hormone-replacement therapy (HRT) has consistently been associa ted with an increased incidence of breast cancer, particularly of small tum ours, Other tumour characteristics in relation to HRT have received less sc ientific attention. Our aim in this population-based prospective cohort stu dy was to assess whether HRT is associated with an increased incidence of b reast-cancer subgroups defined in terms of stage, type (according to the WH O system), Nottingham grade and the Nottingham Prognostic Index (NPI). Eval uation was based on a cohort of 5,865 post-menopausal women followed for an average of 9.8 years. Twenty percent of women reported current use of HRT at the time of the baseline interview. Record linkage with the Swedish Canc er Registry and local clinical registries identified 141 incident invasive breast-cancer cases. All tumours were reclassified by I pathologist, The in cidence of breast cancer in HRT users was 377/10(5) and in non-users 221/10 (5) person-years [relative risk (RR) = 1.72, 95% confidence interval (CI) 1 .17-2.52]. This risk remained statistically significant after adjustment fo r established risk factors in a Cox proportional hazards analysis (RR = 1.6 6, 95% Cl 1.12-2.45), Among HRT users, there was over-representation of cas es with stage I tumours (adjusted RR = 2.33, 95% CI 1.44-3.76), of lobular carcinomas (RR 4.38, 95% CI 1.60-12.0) and of tubular tumours (RR = 4.81, 9 5% CI 1.37-16.8), Nottingham grade I/II carcinomas (RR = 2.02, 95% CI 1.29- 3.16) and cases with NPI less than or equal to 3.4 (RR = 2.29, 95% CI 1.41- 3.72) were similarly over-represented among HRT users. Incidence of breast cancer was increased in postmenopausal women who used HRT at baseline. Amon g HRT users, there was over-representation of tumours that, with regard to stage, type and grade, are associated with a favourable prognosis. (C) 2001 Wiley-Liss. Inc.