The Epidemiology of Early Life Factors and Breast Cancer Risk by Molecular Subtype

Breast cancer can be divided into molecular subtypes, and we have chosen to categorize the disease into six groups. It has been suggested that factors that operate in utero and determine birth characteristics, such as birth weight, birth length and placental weight, are associated with breast cancer risk in adult life. In this project, we want to study whether these early life factors are associated with risk for breast cancer later in life, and whether these factors are differentially associated with different subtypes of breast cancer. This project is part of a larger project, and the subtyping of breast cancer in this specific project has taken more time than originally anticipated. Although the subtyping is going on, and will shortly be completed, the PhD student has completed an analysis that was not described in the project description. That analysis used similar data on breast cancer subtypes that has already been completed, and the paper has recently been published in Cancer Epidemiology Biomarkers and Prevention, with the title "Molecular subtypes of breast cancer: long-term incidence trends and prognostic differences". The paper describes trends in breast cancer incidence according to molecular subtypes in Norwegian women born between 1887 and 1977, and in addition to incidence trends, it also includes an analysis of breast cancer survivial - according to molecular subtypes - where cases diagnosed before 1995 are compared with cases diagnosed in 1995 or later. The results remarkable differences in incidence over time between the subtypes; with strong increases in luminal subtypes and very moderate increases in the incidence of non-luminal subtypes. With respect to prognosis, there has been a consistent improvement in breast cancer survival since 1995, and this improvement is observed for all subtypes of breast cancer.

Molecular subtypes of breast cancer are typically classified as luminal A or B, HER-2 receptor positive and triple negative breast cancer. Adult risk factors, including age at menarche, age at first birth, parity, and body mass, have shown differential as sociations with these subtypes, and support a heterogeneous nature of the disease. The intrauterine environment is important for breast cancer development, but perinatal factors, including placental weight and birth size, have not been studied in relation to molecular subtypes. Also, birth size is closely linked to maternal size, and maternal height may be associated with breast cancer risk among daughters, possibly mediated by perinatal factors. Maternal height may also be an effect modifier of the assoc iation of perinatal factors with breast cancer risk. Three specific hypotheses will be investigated: (1) Birth size (weight and length) is differentially associated with risk for different subtypes of breast cancer: we hypothesize that birth size is p ositively associated with luminal A and B breast cancer. (2) Placental weight, an indicator of placental function, is more strongly associated with breast cancer risk than birth size. Associations of placental weight are strongest for the luminal A and B subtypes. (3) Maternal height is positively associated with breast cancer risk in daughters, and the positive associations of placental weight and birth size are most evident in daughters whose mothers were relatively tall. This research will be based on a follow up of 23,478 Norwegian women (born 1920-1966), and will include at least 800 incident breast cancers (1961-2010). Molecular subtypes will be determined using tissue microarray technology applied to stored breast tissue. The results will advan ce our understanding of early roots of breast cancer, and be useful for developing specific prevention strategies for breast cancer, possibly aimed at particular subtypes.