For patients with high-risk breast cancer, the standard of care is neoadjuvant chemotherapy (NAC), but at least 40% do not achieve a pathological complete response (pCR), depending on their tumor type. Ultrasound localization microscopy (ULM) provides super-resolution images of the microvasculature, showing great promise in identifying early vascular changes in response to therapy. To evaluate its potential for therapy monitoring and preselection, we conducted a prospective study in which we examined 20 breast cancer patients using contrast-enhanced ultrasound (CEUS) before their first, second, and fourth NAC cycles. ULM images were obtained by localizing and tracking microbubbles in the motion-corrected CEUS videos. A radiomics analysis revealed that patients achieving pCR showed higher vessel coverage, a more homogeneous vascular network, and differences in various texture features already before therapy. These differences enabled us to classify patients’ therapy responses with 95% accuracy using logistic regression, which was not feasible with clinical reference methods as tumor size measurement, B-mode ultrasound, CEUS, and histology. Furthermore, pCR patients exhibited significantly higher changes in vessel coverage in response to NAC compared to non-pCR patients. These results emphasize the potential of ULM to identify incomplete therapy responses early on, enabling treatment adaptation and improving therapy outcomes for patients.
Porte et al. (Wed,) studied this question.