Abstract
BACKGROUND: Although breast cancer is rare in young women under 30 years of age, its incidence has been increasing among younger populations in recent years. Identifying the immunohistochemical molecular subtype of early-stage noninvasive breast cancer is thus crucial for optimizing therapeutic strategies. This study aimed to investigate the correlation between multimodal imaging features-including ultrasound (US), digital mammography (DM), and digital breast tomosynthesis (DBT)-and immunohistochemical molecular subtypes in young patients with breast cancer (≤30 years) and middle-aged patients with breast cancer (45-55 years). METHODS: A retrospective analysis included 146 young (≤30 years) and 292 middle-aged (45-55 years) patients with breast cancer diagnosed between January 2015 and March 2025. Imaging characteristics were assessed according to the Breast Imaging Report and Data System (BI-RADS) fifth edition. Mass lesions were evaluated for shape, margin, size, breast density, posterior features (on US), and vascularity (on US). Calcified lesions were analyzed for shape and distribution. Clinical and pathological parameters included age, symptoms, immunohistochemical classifications, T stage, histological grade, and lymph node metastasis status. RESULTS: The molecular subtype distribution in young and middle-aged patients, respectively, was as follows: luminal B, 70 and 140 cases; triple negative (TN), 32 and 64 cases; human epidermal growth factor receptor 2 (HER2) overexpression, 25 and 50 cases; and luminal A, 19 and 38 cases. (I) Younger patients demonstrated higher rates of nipple discharge (P=0.007), tumors >5.0 cm (P=0.009), stage T3 lesions (P<0.001), and lymph node metastasis (P<0.001), whereas middle-aged patients had higher rates of more palpable masses, tumors ≤2.0 cm, and stage T1 lesions. (II) US analysis indicated that luminal B tumors in younger patients had increased calcified masses (P=0.004), indistinct margins (P=0.004), and enhanced posterior features (P<0.001), which was in contrast to the noncalcified masses, spiculated margins, and shadowing found in middle-aged patients. TN tumors tended to be round/oval morphologies in younger patients (P=0.008). And TN tumors tended to appear as noncalcified masses in middle-aged patients (P=0.046). (III) DBT analysis indicated that younger patients with luminal B tumors were more likely to have isodense lesions (P=0.008) and microlobulated margins (P=0.003), while middle-aged patients tended to exhibit hyperdense lesions and indistinct margins; TN tumors in younger patients tended to have oval/round morphologies (P<0.001) and microlobulated margins (P=0.03); HER2-overexpressed tumors in younger patients tended to show isolated calcifications (P=0.02) and fine linear or branching calcifications, while HER2-overexpressed tumors in middle-aged patients were more likely to be pleomorphic or coarse calcifications (P=0.006). CONCLUSIONS: Age-specific imaging patterns correlate with molecular subtypes and can predict aggressive subtypes in younger patients with breast cancer. These biomarkers may help guide preoperative planning and targeted therapy.