We propose to test whether HiSS MRI estimates breast cancer risk more accurately than conventional fat- suppressed MRI and X-ray mammography. Reliable measurements of breast density, which is one of the strongest independent risk factors for breast cancer, could significantly improve management of risk and preventive treatment. X-ray mammography predicts increased cancer risk on a population basis but has limited value for guiding treatment of individual patients, due to errors in separation of parenchyma and fat, and errors in measurement of the parenchymal volume. MRI allows 3D imaging of breast parenchyma and detects independent signals from water and fat. MRI detects significant decreases in breast density due to tamoxifen and raloxifen. However, conventional MRI requires suppression of signals from fat in the breast. Fat suppression is sometimes non-uniform and fat suppression pulses affect the water resonance, reducing the accuracy of parenchymal volume measurements. Therefore we propose an innovative approach to measurement of breast cancer risk: We will use high spectral and spatial resolution MR imaging (HiSS MRI) to reliably separate water and fat signals, and to quantitatively measure parenchymal density and changes in breast density due to tamoxifen. We hypothesize that use of quantitative HiSS imaging will reduce variability, standardize measurements, and increase sensitivity to variations in parenchymal volume. In addition, HiSS may provide new markers for risk based on the novel computer-aided diagnosis (CAD) analysis of the enhanced texture of HiSS images, and HiSS images of fat distribution. This proposal builds on work in this laboratory that implemented HiSS MRI on clinical magnets and demonstrated significant advantages of this approach. We propose the following Specific Aims: Aim #1: Implement quantitative full bilateral measurements of breast parenchymal density and morphology with HiSS MRI. Phantoms attached to the breast will be used to accurately measure MRI- detectable proton density. Novel software will evaluate parenchymal volume and texture. Fat distribution and composition will be tested as additional markers for risk. Aim #2. Compare HiSS with conventional approaches: We will compare breast density measured with HiSS, conventional fat-suppressed MRI, Dixon MRI methods, and X-ray mammography. Measures of breast density will include absolute and fractional parenchymal volume, and MRI-detectable water proton density. The variability of HiSS and conventional MRI measurements will be compared. Aim #3. Measure changes in parenchymal density due to selective estrogen receptor modulator (SERM) therapy: As a measure of the sensitivity of HiSS MRI to changes in breast cancer risk, we propose to image response to preventive therapy that is known to reduce risk. We will compare the sensitivity of HiSS MRI to effects of tamoxifen therapy with that of 2D X-ray mammography and conventional MRI.

R01CA167785

2013-02-08

2020-01-31