S Ciatto1, R Bonardi, M Zappa. 1. Centro per lo Studio e la Prevenzione Oncologica, Viale A. Volta 171, 50131 Firenze FI. careggi@tin.it
Abstract
PURPOSE: To analyse the association of hormone replacement therapy (HRT) and radiological density, defined as the fraction of breast volume occupied by fibroglandular opacity. MATERIAL AND METHODS: We considered 678 postmenopausal women, consecutively undergoing mammography within an organized screening program, 188 (27.7%) of whom were receiving HRT. Attribution of radiological density was performed prospectively and blind as to HRT status. RESULTS: Both radiological density (p = 0.12) and exposure to HRT (p < 0.001) were found to correlate with age (younger women are more likely to use HRT and to have denser breasts, even without HRT) so the analysis of the correlation between HRT and radiological density was adjusted for the possible confounding effect of age. After stratification by age group, univariate analysis showed a significant correlation between HRT and radiological density (density = > 50%: HRT+ = 34.0%, HRT- = 17.6%: F-value 18.86, df = 1, p < 0.001). Also the duration of HRT was significantly associated to radiological density (Pearson coeff. = 0.12: p = 0.0011). Multivariate analysis adjusting for the confounding effect of age showed an increased risk of dense breast (17% or 7% for a radiological density > 25% or > 50%, respectively) in HRT users with respect to non users. DISCUSSION: The study design, the adoption of a classification of breast density less exposed to intraobserver variability and multivariate analysis of a large series, allows a reliable evaluation of the association of HRT to breast density. As no threshold value is known for the masking effect of density on breast cancer, the negative effect of HRT on screening sensitivity cannot be accurately quantified. CONCLUSIONS: This study confirms the effect of HRT on radiological density in postmenopausal women, suggesting a possible reduction of screening sensitivity and efficacy. Such a hypothesis still needs validation by means of a prospective analysis of the correlation between HRT and the frequency of interval cancer occurrence (the best indicator of sensitivity).
PURPOSE: To analyse the association of hormone replacement therapy (HRT) and radiological density, defined as the fraction of breast volume occupied by fibroglandular opacity. MATERIAL AND METHODS: We considered 678 postmenopausal women, consecutively undergoing mammography within an organized screening program, 188 (27.7%) of whom were receiving HRT. Attribution of radiological density was performed prospectively and blind as to HRT status. RESULTS: Both radiological density (p = 0.12) and exposure to HRT (p < 0.001) were found to correlate with age (younger women are more likely to use HRT and to have denser breasts, even without HRT) so the analysis of the correlation between HRT and radiological density was adjusted for the possible confounding effect of age. After stratification by age group, univariate analysis showed a significant correlation between HRT and radiological density (density = > 50%: HRT+ = 34.0%, HRT- = 17.6%: F-value 18.86, df = 1, p < 0.001). Also the duration of HRT was significantly associated to radiological density (Pearson coeff. = 0.12: p = 0.0011). Multivariate analysis adjusting for the confounding effect of age showed an increased risk of dense breast (17% or 7% for a radiological density > 25% or > 50%, respectively) in HRT users with respect to non users. DISCUSSION: The study design, the adoption of a classification of breast density less exposed to intraobserver variability and multivariate analysis of a large series, allows a reliable evaluation of the association of HRT to breast density. As no threshold value is known for the masking effect of density on breast cancer, the negative effect of HRT on screening sensitivity cannot be accurately quantified. CONCLUSIONS: This study confirms the effect of HRT on radiological density in postmenopausal women, suggesting a possible reduction of screening sensitivity and efficacy. Such a hypothesis still needs validation by means of a prospective analysis of the correlation between HRT and the frequency of interval cancer occurrence (the best indicator of sensitivity).
Authors: D Bernardi; M Pellegrini; S Di Michele; P Tuttobene; C Fantò; M Valentini; M Gentilini; S Ciatto Journal: Radiol Med Date: 2012-01-07 Impact factor: 3.469