Abstract
The classification of breast neuroendocrine neoplasms (Br-NENs) was modified many times over the years and is still a matter of discussion. In the present study, we aimed to evaluate the diagnostic reproducibility and impact on patient outcomes of the most recent WHO 2019 edition of breast tumor classification, namely, for neuroendocrine tumors (NETs) and neuroendocrine carcinomas (NECs). This multicentric observational study included 287 breast neoplasms with NE differentiation. The cases were blindly classified by three independent groups of dedicated breast and/or endocrine pathologists following the 2019 guidelines. Diagnostic concordance and clinical impact were assessed. We observed only a moderate overall diagnostic agreement across the three centers (Cohen's kappa 0.4532) in distinguishing NET from solid papillary carcinomas (SPCs) and no special type carcinomas (NST) with NE differentiation. Br-NENs were diagnosed in 122/287 (42.5%) cases, subclassified as 11 NET G1 (3.8%), 84 NET G2 (29.3%), and 27 NEC (9.4%), the latter group consisting of 26 large-cell and 1 small-cell NECs. The remaining 165/287 (57.5%) cases were labeled as non-NEN, including SPC, mucinous, NST, and mixed NE carcinomas. While NET and non-NEN cases had a comparable outcome, the diagnosis of NECs showed negative impact on disease-free interval compared to NETs and non-NENs (p = 0.0109). In conclusion, the current diagnostic classification of Br-NENs needs further adjustments regarding morphological and immunohistochemical criteria to increase the diagnostic reproducibility among pathologists. Our data suggest that, apart from high-grade small- and large-cell NECs, Br-NENs behave like non-NEN breast carcinomas and should be managed similarly.