Grade III Ductal Carcinoma In Situ Arising Within a Long-Standing Fibroadenoma: A Case Report Illustrating the Limits of Shape-Based Breast Imaging Reporting and Data System (BI-RADS) and AI-Assisted Ultrasound Assessment.

A 40-year-old woman with no personal or family history of breast cancer had a right breast mass first characterized as a probable fibroadenoma on an outside MRI evaluation. A repeat MRI performed years later again described a well-circumscribed, benign-pattern (type 1) enhancing nodule assessed as Breast Imaging Reporting and Data System (BI-RADS) 3. Targeted ultrasound showed an oval, parallel-oriented, circumscribed hypoechoic mass, and an integrated artificial intelligence (AI) lesion-characterization tool returned a categorical output of benign (BI-RADS 2) based on the shape and orientation descriptors displayed on screen. However, shear-wave elastography of the same lesion showed markedly elevated stiffness (maximum elasticity 135.48 kPa), prompting the radiologist to revise the overall assessment to BI-RADS 4 and pursue ultrasound-guided tru-cut biopsy, which revealed solid, cribriform, and comedo-type high-grade (nuclear grade 3) ductal carcinoma in situ (DCIS) arising within a fibroadenoma. Whole-body 18F-fluorodeoxyglucose (18F-FDG) PET/CT subsequently showed FDG uptake at the known primary site with no other avid disease, while histopathology established the diagnosis and stage. The patient underwent segmental mastectomy with sentinel lymph node biopsy; final pathology confirmed a 1.5 × 1 cm focus of high-grade DCIS entirely confined to the fibroadenoma, with clear surgical margins and no nodal involvement (pTisN0(i-,sn)). This case demonstrates that malignancy, including high-grade DCIS, can arise within an otherwise typical, long-stable fibroadenoma and can retain a benign morphological profile on B-mode ultrasound and AI-assisted shape-based classification. Quantitative elastographic stiffness discordant with a benign shape classification should not be dismissed and warrants biopsy. Multiparametric correlation, combining morphologic assessment with quantitative elastography, remains essential and should not be superseded by a single shape-based descriptor, human or algorithmic. Ultimately, in this case, only tissue sampling provided a clear and definitive diagnosis, underscoring that biopsy should not be deferred when a breast mass is reassessed after a long interval or shows any feature discordant with a straightforward fibroadenoma.