BACKGROUND: A premalignant lesion in the breast is associated with an increased risk of breast cancer. The aim of this article was to identify women with an increased risk of breast cancer based on prior screening results (PSRs). METHODS: This registry-based cohort study followed women who participated in the organized breast cancer screening program in Norway, BreastScreen Norway, in 1995-2016. Incidence rates and incidence rate ratios were used to estimate absolute and relative risks of breast cancer associated with PSRs. Histopathological characteristics of subsequent breast cancers were presented by PSRs. RESULTS: This study included 762,643 women with up to 21 years of follow-up. In comparison with negatively screened women, increased incidence rate ratios of 1.8, 2.0, 2.9, and 3.8 were observed after negative additional imaging, for benign biopsy, for hyperplasia with atypia, and for carcinoma in situ, respectively. Subsequent breast cancers did not differ in tumor diameter or histological grade, whereas the proportion of lymph node-positive breast cancers decreased as the presumed malignancy potential of PSRs increased. CONCLUSIONS: The risk of subsequent breast cancer increased with the presumed malignancy potential of PSRs, whereas the tumor characteristics of subsequent cancers did not differ except for the lymph node status. Women with screen-detected benign lesions or hyperplasia with atypia might benefit from more frequent screening. Cancer 2019;125:3330-3337.
Physical inactivity, high postmenopausal body mass index, alcohol consumption and use of menopausal hormone therapy are established risk factors for breast cancer.Less is known about whether these factors influence the risk of progression of benign and premalignant breast lesions to invasive breast cancer. This registry-based cohort study was based on women with a precancerous lesion who were followed for breast cancer. The cohort consisted of 11 270 women with a benign lesion, 972 women with hyperplasia with atypia and 2379 women with carcinoma in situ diagnosed and treated after participation in BreastScreen Norway, 2006-2016. Information on breast cancer risk factors was collected by a questionnaire administered with the invitation letter. Cox regression analysis was used to estimate the association between breast cancer and physical activity, body mass index, alcohol consumption, tobacco smoking and menopausal hormone therapy, adjusted for age. During follow-up, 274 women with a benign lesion, 34 women with hyperplasia with atypia and 118 women with carcinoma in situ were diagnosed with invasive breast cancer.We observed an increased risk of breast cancer associated with use of menopausal hormone therapy for women with a benign or premalignant lesion. Alcohol consumption and tobacco smoking showed suggestive increased risk of breast cancer among women with a benign lesion. We were only to a limited degree able to identify associations between modifiable risk factors of breast cancer and the disease among women with a precancerous lesion, and a larger study is needed to confirm or refute associations.
This paper construct upper and lower bounds for expected values of convex functions for Multivariate Totally Positive distributions of order 2. The bounds can be iteratively improved, and can be used to optimize information measures or value of information utilities for Bayesian Networks with the property of being Multivariate Totally Positive of order 2. The bounds are applied on a simple illustrating synthetic example with binary variables for simplicity.The construction of the bounds is inspired by the local updates made by the message passing in the Junction Tree Algorithm, and utilizes the coherent covariance structure of Multivariate Totally Positive distributions of order 2. The resulting formula allows for preprocessing in the full joint to construct pairs of smaller local distributions covering different parts of the network, from which the upper and lower bounds are efficiently constructed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.