Dedicated breast PET and MRI for characterization of breast cancer and its response to therapy

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT The objective of this academic-industrial partnership (AIP) project is to demonstrate the utility of dedicated breast positron emission tomography (dbPET) for characterizing primary breast cancers and their response to neoadjuvant chemotherapy (NAC). While dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) depicts changes in tumor morphology and vascularity in response to NAC, dbPET provides complementary information about tumor metabolism that can powerfully predict treatment response earlier in the course of therapy. We focus this project on MAMMI dbPET by OncoVision because it provides the crucial combination of high spatial resolution and sensitivity that can enable accurate intratumoral mapping of metabolic changes in small lesions with only half the radiotracer dose of whole-body PET. Importantly, this new system scales PET technology to be both economically and clinically feasible in the early (pre-metastatic) breast cancer setting. As the imaging lead (PI: Nola Hylton) of the I-SPY 2 TRIAL, a clinical trial designed to identify novel therapeutics for breast cancer, we are in a unique position to integrate and test the performance of FDG-dbPET as an early marker for treatment response. As academic-industrial partners, UCSF and OncoVision will work together to develop a user-friendly and cost-effective dbPET technology that can be easily adopted into the clinical workflow of most breast cancer centers. In Specific Aim 1, we will develop software capabilities to standardize dbPET image registration and quantification to accurately quantify longitudinal changes with treatment. In Specific Aim 2, we will acquire pre- and post-treatment FDG-dbPET images of a subset of I-SPY 2 patients and clinically evaluate whether tumor metabolic metrics (i.e., optimized standardized uptake values, SUV) from dbPET can act as early predictors of pathologic complete response — in comparison to, and in combination with, the functional tumor volume (FTV) metric from DCE-MRI. We will test the biomarker performance of dbPET SUV and combined SUV+FTV using logistic regression predictive models. We will also explore the association of dbPET and DCE-MRI radiomic features with breast cancer biomarkers in order to identify imaging features with prognostic value. In addition, our prospective study’s dbPET data will be used to evaluate the software capabilities developed in Specific Aim 1. We expect the successful completion of this AIP project to enable the use of MAMMI dbPET in routine breast cancer management and to produce a set of imaging biomarkers relevant to tumor biology and its change in response to treatment.