Relationships between computer-extracted mammographic texture pattern features and BRCA1/2 mutation status: a cross-sectional study.

INTRODUCTION: Mammographic density is similar among women at risk of either sporadic or BRCA1/2-related breast cancer. It has been suggested that digitized mammographic images contain computer-extractable information within the parenchymal pattern, which may contribute to distinguishing between BRCA1/2 mutation carriers and non-carriers. METHODS: We compared mammographic texture pattern features in digitized mammograms from women with deleterious BRCA1/2 mutations (n = 137) versus non-carriers (n = 100). Subjects were stratified into training (107 carriers, 70 non-carriers) and testing (30 carriers, 30 non-carriers) datasets. Masked to mutation status, texture features were extracted from a retro-areolar region-of-interest in each subject's digitized mammogram. Stepwise linear regression analysis of the training dataset identified variables to be included in a radiographic texture analysis (RTA) classifier model aimed at distinguishing BRCA1/2 carriers from non-carriers. The selected features were combined using a Bayesian Artificial Neural Network (BANN) algorithm, which produced a probability score rating the likelihood of each subject's belonging to the mutation-positive group. These probability scores were evaluated in the independent testing dataset to determine whether their distribution differed between BRCA1/2 mutation carriers and non-carriers. A receiver operating characteristic analysis was performed to estimate the model's discriminatory capacity. RESULTS: In the testing dataset, a one standard deviation (SD) increase in the probability score from the BANN-trained classifier was associated with a two-fold increase in the odds of predicting BRCA1/2 mutation status: unadjusted odds ratio (OR) = 2.00, 95% confidence interval (CI): 1.59, 2.51, P = 0.02; age-adjusted OR = 1.93, 95% CI: 1.53, 2.42, P = 0.03. Additional adjustment for percent mammographic density did little to change the OR. The area under the curve for the BANN-trained classifier to distinguish between BRCA1/2 mutation carriers and non-carriers was 0.68 for features alone and 0.72 for the features plus percent mammographic density. CONCLUSIONS: Our findings suggest that, unlike percent mammographic density, computer-extracted mammographic texture pattern features are associated with carrying BRCA1/2 mutations. Although still at an early stage, our novel RTA classifier has potential for improving mammographic image interpretation by permitting real-time risk stratification among women undergoing screening mammography.

Mammographic density does not differ between unaffected BRCA1/2 mutation carriers and women at low-to-average risk of breast cancer.

Elevated mammographic density (MD) is one of the strongest risk factors for sporadic breast cancer. Epidemiologic evidence suggests that MD is, in part, genetically determined; however, the relationship between MD and BRCA1/2 mutation status is equivocal. We compared MD in unaffected BRCA1/2 mutation carriers enrolled in the U.S. National Cancer Institute's Clinical Genetics Branch's Breast Imaging Study (n = 143) with women at low-to-average breast cancer risk enrolled in the same study (n = 29) or the NCI/National Naval Medical Center's Susceptibility to Breast Cancer Study (n = 90). The latter were BRCA mutation-negative members of mutation-positive families or women with no prior breast cancer, a Pedigree Assessment Tool score <8 (i.e., low risk of a hereditary breast cancer syndrome) and a Gail score <1.67. A single experienced mammographer measured MD using a computer-assisted thresholding method. We collected standard breast cancer risk factor information in both studies. Unadjusted mean percent MD was higher in women with BRCA1/2 mutations compared with women at low-to-average breast cancer risk (37.3% vs. 33.4%; P = 0.04), but these differences disappeared after adjusting for age and body mass index (34.9% vs. 36.3%; P = 0.43). We explored age at menarche, nulliparity, age at first birth, menopausal status, number of breast biopsies, and exposure to exogenous hormonal agents as potential confounders of the MD and BRCA1/2 association. Taking these factors into account did not significantly alter the results of the age/body mass index-adjusted analysis. Our results do not provide support for an independent effect of BRCA1/2 mutation status on mammographic density.

Effect of raloxifene on mammographic density and breast magnetic resonance imaging in premenopausal women at increased risk for breast cancer.

BACKGROUND: Mammographic density is a risk factor for breast cancer. Mammographic density and breast magnetic resonance imaging (MRI) volume (MRIV) assess the amount of fibroglandular tissue in the breast. Mammographic density and MRIV can be modulated with hormonal interventions, suggesting that these imaging modalities may be useful as surrogate endpoint biomarkers for breast cancer chemoprevention trials. We evaluated the effect of raloxifene on mammographic density and MRIV in premenopausal women at increased risk for breast cancer. METHODS: Mammograms and MRI were obtained at baseline and after 1 and 2 years of 60 mg raloxifene by mouth daily for 27 premenopausal women. Mammographic percent dense area was calculated using a semiquantitative thresholding technique. T(1)-weighted spoiled gradient-echo MRI with fat suppression was used to determine breast MRIV using a semiautomatic method. Mean change in mammographic density and median change in MRIV were assessed by the Wilcoxon signed-rank test. RESULTS: No significant change in mammographic density was seen after treatment with raloxifene. Mean change after 1 year was 1% [95% confidence interval (95% CI), -3 to +5] and after 2 years was 1% (95% CI, -2 to +5). MRIV decreased on raloxifene. Median relative change in MRIV after 1 year was -17% (95% CI, -28 to -9; P = 0.0017) and after 2 years was -16% (95% CI, -31 to -4; P = 0.0004). CONCLUSIONS: In high-risk premenopausal women, mammographic density did not change on raloxifene, whereas MRIV significantly declined. Our findings suggest that MRIV is a promising surrogate biomarker in premenopausal women at increased risk for breast cancer and should be investigated further in breast cancer prevention trials.

Exemestane Use in Postmenopausal Women at High Risk for Invasive Breast Cancer: Evaluating Biomarkers of Efficacy and Safety.

This phase II trial evaluated clinical markers of efficacy and safety of exemestane in postmenopausal women at increased risk for breast cancer. Postmenopausal women (n = 42) at risk for invasive breast cancer received 25 mg exemestane daily for 2 years along with calcium and vitamin D. The primary outcome was change in mammographic density (MD) after one year. Secondary outcomes included change in serum steroid hormones as well as change in trefoil protein 1 (TFF1) and proliferating cell nuclear antigen (PCNA) in breast tissue. Safety and tolerability were also assessed. MD decreased at 1 year and was significant at 2 years [mean change = -4.1%; 95% confidence intervals (CI), -7.2 to -1.1; P = 0.009]. Serum estradiol and testosterone levels significantly decreased at 3 months and remained suppressed at 12 months. After 1 year of treatment, TFF1 intensity decreased (mean change -1.32; 95% CI, -1.87 to -0.76; P < 0.001). Exemestane was safe and well tolerated. Exemestane decreased MD and expression of breast tissue TFF1. It was well tolerated with few clinically relevant side effects. MD and breast tissue TFF1 are potential biomarkers of breast cancer-preventive effects of exemestane in high-risk postmenopausal women.