An observational study to assess treatment success in HER2-positive breast cancer prior to surgery using advanced imaging (18F-FDG PET/MRI) and AI-assisted analysis

HER2-positive breast cancer is a particularly aggressive form of breast cancer in which the HER2 protein is overexpressed on cancer cells, leading to faster tumor growth. As a standard practice, patients with HER2-positive breast cancer receive neoadjuvant therapy—a combination of chemotherapy and antibody therapy—before surgery to shrink the tumor.

As a standard practice, staging is performed via a CT scan prior to treatment. During and after the preoperative drug treatment, further imaging (usually ultrasound, breast MRI, or CT scanning) is used to assess how well the tumor has responded to therapy. The imaging determines whether the tumor has shrunk or completely regressed. To date, the final assessment of the treatment response has been determined exclusively through histopathological examination of breast tissue and affected lymph nodes following surgery. If the tumor has completely regressed, this is referred to as a pathological complete response (pCR), which is associated with a better prognosis. To date, there is no imaging method that can reliably predict whether a pCR is present. In this study, the imaging method used is 18F-FDG PET/MRI. This method combines two techniques: positron emission tomography (PET) and magnetic resonance imaging (MRI).

18F-FDG PET/MRI is an established imaging method already used in breast cancer diagnosis and treatment monitoring. PET shows how active metabolic processes are in the body. To do this, a weakly radioactive sugar (in this case, 18F-FDG) is administered intravenously. Cancer cells absorb this sugar particularly rapidly because they consume more energy than healthy cells. This makes it possible to visualize tumors and their activity. MRI provides a very precise view of the structure and tissue of organs and tumors without exposing the patient to radiation. The use of artificial intelligence (AI) to analyze the images is not yet standard practice. AI could recognize specific patterns in the image data, thereby enabling a more precise prediction of whether the tumor has completely regressed even before surgery takes place.

The goal of this observational study is to use precise imaging via 18F-FDG PET/MRI and supported by artificial intelligence (AI), to predict the extent of the disease—and thus the response to and success of therapy—early and reliably. To this end, patients undergo a total of three 18F-FDG PET/MRI scans. The first scan takes place before neoadjuvant treatment and simultaneously replaces the CT scan otherwise recommended as standard for staging. The second scan is performed after the first two cycles, and the third after the completion of neoadjuvant treatment, immediately before surgery. The tissue samples collected during surgery serve as a benchmark for evaluating the imaging results. The primary endpoint is the accuracy of 18F-FDG PET/MRI imaging combined with AI. Participants will be followed for approximately 15 weeks as part of the study.

Eligible participants are women aged 18 and older with HER2-positive breast cancer who are scheduled to undergo neoadjuvant systemic therapy and who do not have tumor spread to other organs (distant metastases). An MRI scan must be medically feasible.