Correlation of breast cancer biology and x-ray scatter data

Abstract

Wide angle x-ray scatter (WAXS) imaging can potentially be used to diagnose ductal carcinoma in situ (DCIS) in breast biopsies. In this work, WAXS signals are used to model healthy and diseased breast tissue. It is accomplished by analysis of the breast tissue’s differential linear scattering coefficients ‘μₛ’. The tissue samples are assumed to consist of varying amount of fat, stroma and cells and their μₛ coefficients can be expressed by the following sum: μₛ(x) = afat • μₛ fat(x) + astr •μₛ str(x) + acell •μₛ cell(x), where x= (1/ λ) sin(θ/2) is the momentum transfer argument, λ is the x-ray wavelength and θ is the scatter angle. Data for breast tissues was taken from literature and that for stroma was estimated using histology studies. The fractional volumes of each of the three components (afat, astr, acell) were evaluated using singular value decomposition (SVD). Fitting results suggest cancer consists of 44% cells as compared to only 8% and 7% for normal and benign tissue, respectively. These results support the argument that an uncontrolled proliferation of cells is the hallmark of cancer. The fat decreases from 56% (normal) to 26% (benign) and to only 3.4% in cancer. This is justified by the fact that invasive ductal carcinoma cells use fat to fuel their growth and spread. Stromal increase from 35% in normal to 67% in benign suggest that benign diseases such as fibroadenoma are characterized by a surge in fibrous tissue. The model predictions of the compositions for normal, benign and malignant tissue are promising. It gives insight into how cancer interacts with its microenvironment. While scatter studies on cells and fat have been done in the past, not much is known about stroma in the same context. Studies show that stroma promotes cancer growth and metastasis, making it an important component to focus on. The second part of this research works on this aspect. A model for stroma made up of 80% collagen, 10% elastin and 10% versican is introduced. μs for stroma is then determined using Independent Atomic Model (IAM) approximation. The scattersignal obtained via this newly introduced stroma model matches with the previous estimation, mainly at the higher values of momentum transfer values. This is because IAM fails at the lower regime due to interference phenomenon. It is observed that the scatter values obtained using IAM are not very sensitive to the change in composition of stroma. In most of the cases, the percentage change with reference to the original model is less than 1%. In the event in which elastin is considered to be the major constituent, a difference of 2.41% to 3.49% is seen. This model gives important information about the structure and function of stroma and highlights its role in cancer progression.