MR and radiomics used to predict breast cancer recurrence

Principal investigator Despina Kontos, PhD, Associate Professor of Radiology in the Perelman School of Medicine at the University of Pennsylvania, stated, "If we’re only taking out a little piece of a tissue from one part of a tumor, that does not give the full picture of a person’s disease and of his or her response to specific therapies." She and her colleagues set out to determine whether they could use MR and radiomics for more personalized tumor characterization.

To do this, they extracted 60 biomarkers from 95 women with primary invasive breast cancer. After following up with the patients 10 years later, the group found that a scan that showed high tumor heterogeneity at the time of diagnosis could successfully predict a cancer recurrence. "Women who had more heterogeneous tumors tended to have a greater risk of tumor recurrence," said the study’s lead author Rhea Chitalia, a PhD candidate in the School of Engineering and Applied Science at the University of Pennsylvania.

The researchers retrospectively analyzed patient scans from a clinical trial conducted at Penn Medicine. For each woman, the group generated a "signal enhancement ratio" map and, from it, extracted imaging features to see the relationship between them and conventional biomarkers (such as gene mutations or hormone receptor status) and patient outcomes. They found their algorithm was able to successfully predict recurrence-free survival after 10 years. To validate their findings, the group compared their results to an independent sample of 163 patients with breast cancer from the publicly available Cancer Imaging Archive.