This Pictorial Essay reviews the imaging and treatment of four main breast cancer molecular subtypes characterized by biomarker analysis: luminal A, luminal B including triple-positive breast cancers, nonluminal HER2-enriched, and basallike including triple-negative breast cancers. Overlapping molecular markers make it difficult to categorize up to 6% of breast cancers. Luminal A and luminal B cancers, characterized by estrogen receptor expression, comprise the majority (79%) of all breast cancers. On imaging, luminal cancers appear as irregular masses with spiculated margins. Nonluminal cancers are fast-growing, aggressive cancers. Nonluminal HER2-enriched cancers (4%) show a higher propensity for multifocality and calcifications associated with high-grade ductal carcinoma in situ, whereas nonluminal basallike cancers (11%) are usually unifocal and noncalcified. Knowledge of the molecular features, typical imaging appearances, and treatment paradigms for each subtype helps radiologists serve as better patient advocates and consultants. However, the limitations of this simplified approach must be acknowledged, as tumor heterogeneity leads to overlap. Advances in computer science, including artificial intelligence and radiomics, offer hope for improved diagnosis, classification, prognostication, and prediction of treatment response.
Chung et al. (Thu,) studied this question.