Dear Editor,
The article by Valkovic Zujic et al. (1) addresses whether vacuum-assisted excision (VAE) can safely reduce surgical overtreatment of B3 lesions diagnosed by tomosynthesis-guided vacuum-assisted biopsy (VAB). The study is commendable for its focus on multidisciplinary de-escalation in a field where practice remains heterogeneous (2).
Several methodological aspects, however, deserve attention when interpreting the conclusion that VAE represents a safe alternative to surgical excision (SE) for B3 lesions (1). The comparison between VAE and SE is intrinsically shaped by allocation decisions made by the multidisciplinary team: lesions considered radiologic-pathologic concordant and lower risk were directed toward VAE, whereas lesions with greater concern for underestimation were preferentially referred for surgery (1). This allocation is clinically appropriate, but has important methodological implications: the VAE and SE cohorts are enriched for different baseline oncologic risk profiles and do not represent equivalent treatment arms; the 30.8% upgrade rate observed in the SE group is a direct consequence of this selection strategy (1, 3).
Second, the absence of malignant upgrades in the VAE group should be interpreted with caution, given the limited number of VAE-treated lesions and the shorter mean follow-up compared with that of the surgical cohort (1). In published retrospective series, upgrade after VAE or VAB is uncommon but not negligible, and is consistently influenced by histologic subtype, particularly by the presence of atypia (3, 4). Recent single-centre data on 366 screen-detected B3 microcalcifications further confirm non-zero upgrade rates despite vacuum-assisted sampling (4).
A zero-upgrade rate in a small, highly selected series is therefore reassuring, yet is better viewed as hypothesis-generating than as definitive proof of long-term oncologic safety (1, 3). This interpretation aligns with current guideline-based stratification. The 2024 European recommendations (European Society of Breast Cancer Specialists/European Society of Breast Imaging/European Society of Pathology/European Society of Surgical Oncology) support VAE or surveillance for selected B3a lesions when there is radiologic-pathologic concordance, while recommending surgery for ADH/AIDEP and lobular carcinoma in situ/lobular intraepithelial neoplasia grade 2, whose upgrade rates remain higher (2, 4). In the present study, ADH/AIDEP accounted for most surgical upgrades, reinforcing that VAE’s strength lies in selective use rather than in broad substitution for surgery (1). Prior ultrasound-guided 8-gauge VAE experience with BI-RADS 3–4 masses points in the same direction, with lesion size and imaging features emerging as key determinants of complete excision (5). Although ultrasound-guided and tomosynthesis-guided settings are not directly superimposable, both approaches converge on the same practical message: minimally invasive excision is most reliable when grounded in strict radiologic-pathologic concordance and subtype-specific risk assessment (1, 2, 5).
These observations contextualise rather than diminish the authors’ contribution. The study supports multidisciplinary team-guided de-escalation, but underscores the need for larger prospective registries, standardized subtype-specific reporting, and longer follow-up before equivalence with surgery can be claimed (3). In this evolving field, the challenge is not simply to replace surgery with minimally invasive excision, but to match the intensity of intervention to the biological and radiologic risk of each B3 lesion, so that de-escalation becomes a matter of precision rather than simplification.
