Skip to main content English


Multiparametric 18F-Fluoroestradiol PET/MRI coupled with Radiomics Analysis and Machine Learning for Prediction and Assessment of Response to Neoadjuvant Endocrine Therapy in Patients with Hormone Receptor+/HER2− Invasive Breast Cancer

Katja Pinker-Domenig, MD, PhD

Principal Investigator

Medical University of Vienna & General Hospital
Department of Biomedical Imaging and Image-guided Therapy
Waehringer Guertel 18-20, Floor 7F
1090 Vienna, Austria


Jubiläumsfonds der der Österreichischen Nationalbank Projektnummer # 18207

Overarching challenge and innovative aspects

Breast cancer (BC) is a heterogeneous disease and, as a result, breast tumors respond differently to specific therapies. Studies based on global gene expressions analysis have allowed for the emergence of four major classes of breast cancer: Luminal A, Luminal B, human epidermal growth factor receptor 2 (HER2) enriched, and basal-like. These four intrinsic types of breast cancers differ with respect to treatment recommendations, outcomes, and response to therapies. Further genomic analysis will likely continue to sub-classify breast cancers.

A major challenge in the treatment of the most common type of BC (hormone receptor positive [HR+]/HER2−) is the prospective selection of patients who will derive optimum benefits from a specific therapy or approach (i.e., chemotherapy and/or endocrine therapy). Multigene assays such as Endopredict (Myriad Genetic Laboratories, Inc, Salt Lake City, UT), Oncotype DX (Genomic Health, Redwood City, CA) and MammaPrint (Agendia, Irvine, CA) have proven valuable for prognosis and have been found to predict treatment benefit for women with early-stage breast cancer. These are currently routinely utilized in clinical practice. Nonetheless, since avoiding and minimizing unnecessary toxicities for these patients who will not benefit from a specific therapy is of utmost importance, further research in this field is ongoing.

Herein we propose to identify a platform that will aid in prospectively identifying effective neoadjuvant therapies for the treatment of HR+/HER2− BC. We will undertake the following:

  • Decipher the inter- and intratumoral heterogeneity of the hormone receptor estrogen (ER) using high-resolution multiparametric 16a-18F-fluoro-17b-estradiol (18F-FES) positron emission tomography/magnetic resonance imaging (PET/MRI) with molecular tumor profiles whereby the coupling will be achieved through radiomics analyses and machine learning (ML).
  • Develop a multi-layered PET/MRI approach for quantitative assessment and spatial mapping of functional ER using new MRI techniques and the radiotracer 18F-FES.
  • Identify radiomics signatures predictive of neoadjuvant endocrine response in patients with HR+/HER2− invasive BC.