OBJECTIVES: To determine if breast cancer risk in women with SLE is modified by a history of exposure to hormone replacement therapy (HRT) or oral contraceptives (OC), after adjusting for other risk factors. METHODS: Data were pooled from SLE cohorts at three centres. For each female cohort member (n = 871), the probability of developing breast cancer was estimated from factors (age, parity, age at first live birth, age of menarche, personal history of benign breast disease, family history) in the Gail model, an established tool for predicting breast cancer risk. From these probabilities, the expected number of breast cancers for the cohort was estimated. Actual occurrence of cases was determined by linkage with regional cancer registries. Standardized incidence ratios (SIRs; ratio of cancers observed to expected) were calculated, with subgroup analyses according to HRT and OC exposure. RESULTS: In the cohort, 15 breast cancers occurred vs 7.2 predicted [SIR 2.1, 95% confidence interval (CI) 1.1, 3.5]. When controlling for Gail model risk factors, estimates were similar for women never exposed to HRT vs those exposed to HRT. Adjusted SIR estimates appeared similar also for women exposed or not exposed to OC. CONCLUSIONS: Although not definitive, the data suggest that the breast cancer experience in this sample is not completely explained by factors such as reproductive and family history, or by exogenous hormonal exposures. Other determinants, including medication exposures or genetic factors (possibly related to oestrogen receptors or metabolism) may be important. Variations in these factors might explain why an elevated risk of breast cancer has not been apparent in all SLE populations.
OBJECTIVES: To determine if breast cancer risk in women with SLE is modified by a history of exposure to hormone replacement therapy (HRT) or oral contraceptives (OC), after adjusting for other risk factors. METHODS: Data were pooled from SLE cohorts at three centres. For each female cohort member (n = 871), the probability of developing breast cancer was estimated from factors (age, parity, age at first live birth, age of menarche, personal history of benign breast disease, family history) in the Gail model, an established tool for predicting breast cancer risk. From these probabilities, the expected number of breast cancers for the cohort was estimated. Actual occurrence of cases was determined by linkage with regional cancer registries. Standardized incidence ratios (SIRs; ratio of cancers observed to expected) were calculated, with subgroup analyses according to HRT and OC exposure. RESULTS: In the cohort, 15 breast cancers occurred vs 7.2 predicted [SIR 2.1, 95% confidence interval (CI) 1.1, 3.5]. When controlling for Gail model risk factors, estimates were similar for women never exposed to HRT vs those exposed to HRT. Adjusted SIR estimates appeared similar also for women exposed or not exposed to OC. CONCLUSIONS: Although not definitive, the data suggest that the breast cancer experience in this sample is not completely explained by factors such as reproductive and family history, or by exogenous hormonal exposures. Other determinants, including medication exposures or genetic factors (possibly related to oestrogen receptors or metabolism) may be important. Variations in these factors might explain why an elevated risk of breast cancer has not been apparent in all SLE populations.