Quantitative Measures by Dedicated Breast PET and MRI for Early Prediction of Response to Neoadjuvant Chemotherapy
Abstract
The goal of this project is to develop a non-invasive imaging method to inform whether patients are responding to chemotherapy early in the course of treatment. It will guide treatment redirection to a more effective regimen for patients whose tumors fail to respond but to forego additional chemotherapy for patients with excellent response, sparing them from unnecessary toxic therapy. This project directly addresses the overarching challenges of (1) distinguishing deadly from non-deadly breast cancers and (2) conquering the problems of overdiagnosis and overtreatment. This project seeks to use a high-resolution breast positron emission tomography (PET) with magnetic resonance imaging (MRI) and computational methods to estimate the relationship linking imaging features and treatment outcome. Through computer programming, mathematical models will be built and automated to learn from the imaging data and then make accurate predictions about early response outcome to enable treatment modifications for patients at the extreme ends of the response spectrum. In the pre-operative treatment setting, it is now known that patients whose tumors are completely gone at the time of surgery have excellent survival outcomes. In contrast, patients with substantial disease remaining at the time of surgery have much poorer outcomes. Imaging is useful for monitoring primary tumor response during treatment. In the past, the size of the tumor measured by MRI has shown to be predictive of treatment response. The addition of breast PET, using a radio-sugar (FDG) tracer before and at an early treatment point, will provide specific tumor metabolic information that may further improve the predictive value of imaging. The dedicated breast PET (dbPET) proposed in this project produces detailed images approximately five times better than the whole-body PET (wbPET). Detailed imaging patterns obtained from dbPET may reflect the biology of the tumor microenvironment, which dictates the tumor behavior to treatment response. Thus, in-depth analysis of imaging patterns in the primary breast cancer may lead to a more accurate prediction of response to therapy. The integration of dbPET into breast cancer management therefore represents a departure from the traditional use of wbPET for assessing metastatic disease. Instead, the dbPET technology provides critical molecular insights of primary non-metastatic breast tumors to better guide treatment selection and to better assess early molecular changes in response to treatment. To further explore the significance of imaging patterns in relation to aggressive tumor behaviors, dbPET imaging with a radio-estrogen tracer (FES) will be used to interrogate estrogen receptor positive (ER+) breast cancers. This part of the project is motivated by the fact that ER+ breast cancer patients often have to face the dilemma of whether they should be treated with chemotherapy or with hormone therapy alone. Using the similar image analysis and modeling approach described above, imaging patterns tiding to recurrence risk may emerge as classifiers to guide more effective treatment decision for this common breast cancer subtype. If successful, this work may produce a breakthrough technology that enables digital assessment of the primary breast tumor early in the treatment to guide better clinical decisions to improve breast cancer patients’ quality of life during chemotherapy and their overall survival.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 29, 2018
- Source ID
- W81XWH1810671
Entities
People
- Ella F Jones
Organizations
- United States Army
- University of California, San Francisco